An overall obstacle avoidance method for multi-unmanned vehicle formation based on improved artificial potential field method

Abstract

A multi-unmanned vehicle formation obstacle avoidance method based on an improved artificial potential field method. This invention combines and improves the artificial potential field method with the pilot-follow method. When the pilot detects an obstacle, the method considers the overall formation and calculates the traversable distance based on sensor data. Based on the combined force of the target's attraction, the obstacle's repulsion, and an additional repulsive force, the pilot is directed to the center of the traversable area. The pilot then guides the formation to avoid the obstacle. The advantages of this method are that it combines the improved artificial potential field method with the pilot-follow method, introduces an additional repulsive force to balance the pilot's horizontal force, and rationally utilizes terrain to achieve formation obstacle avoidance. This solves the problem of frequent formation changes in complex environments encountered by most solutions.

Description

Technical Field

[0001] The present invention relates to the field of robotics technology, and in particular to an overall obstacle avoidance method for a multi-unmanned vehicle formation based on an improved artificial potential field method. Background Art

[0002] In recent years, with the continuous development of automatic control, communication, and robotics technologies, the performance and reliability of unmanned vehicles have greatly improved. More and more unmanned vehicles are appearing in production and daily life, continuously providing convenience for humanity. After years of research and development, unmanned vehicle technology has matured, and the number of unmanned vehicles to meet various needs has continued to increase, and the variety of unmanned vehicles has also increased. However, a single unmanned vehicle is generally not capable of performing complex tasks, making targeted development difficult and costly. When performing some complex tasks, a single unmanned vehicle is insufficient, which has led to the emergence of multi-unmanned vehicle systems.

[0003] Currently, research on obstacle avoidance for multi-AV platooning is still immature. Algorithms for platooning obstacle avoidance primarily draw on the traditional obstacle avoidance strategies of individual AVs and apply them to multi-AV platooning. These include fuzzy logic algorithms, artificial potential field methods, and genetic algorithms. Platooning obstacle avoidance requires more sophisticated techniques. In an obstacle-filled environment, the AVs must maintain their formation while also avoiding obstacles appropriately.

[0004] The currently available prior art is as follows:

[0005] Patent number: CN202110194494.1, patent name: Multi-mobile robot formation obstacle avoidance control method based on artificial potential field method, which discloses a multi-mobile robot formation obstacle avoidance control method based on artificial potential field method, including: before detecting an obstacle, multiple mobile robots form an initial formation and move towards the target point; when the formation members detect an obstacle, the overall speed of the multi-mobile robot formation is reduced, and the formation obtains the distance between the obstacles according to the magnitude of the obstacle repulsion, and determines whether the initial formation can pass through the obstacle. Otherwise, the formation is switched to avoid the obstacle, and a reasonable formation is selected from the formation database to avoid the obstacle. When the formation cannot detect the obstacle, the formation successfully avoids the obstacle; the formation is switched to the initial formation, the initial speed is restored, and the target point is reached to complete the task. The advantage of the present invention is that it ensures that the formation as a whole can pass through obstacles quickly, efficiently and reasonably, and can successfully complete obstacle avoidance.

[0006] However, it uses the potential field to determine the width of the obstacle area that can be passed, then compares it with the initial formation width. Based on the comparison results, it formulates an appropriate obstacle avoidance strategy and designs a rapid formation recovery after obstacle avoidance. This solves the problem of unreasonable formation transformation in the presence of obstacles.

[0007] This application proposes an overall obstacle avoidance method for a multi-unmanned vehicle formation based on an improved artificial potential field method. The force analysis is performed on the navigator when there are obstacles on both sides. By introducing additional repulsive force, the navigator is enabled to move to the middle of the passable area under the action of gravity, repulsion, and additional repulsive force, and then move forward from the middle of the passable area until it moves out of the obstacle influence area, providing the largest possible operating area for the two followers and reducing the changes in formation.

[0008] Patent number: CN201911162061.7, patent name: A formation-maintaining method suitable for quadcopter UAV swarm flight. This method determines the parameters of the UAV swarm based on the initial formation when the UAV swarm takes off. During flight, the formation is organized through a formation reference point. After the leader UAV reaches the target point and hovers, the follower UAV can track the formation reference point, eventually forming the desired formation and hovering. Obstacle avoidance and collision prevention between UAVs in the formation are performed through the artificial potential field method. Forces are generated between the target point, the formation reference point, obstacles and UAVs, and between UAVs. This method can effectively ensure that the UAV swarm flies in a preset formation during flight. At the same time, it can avoid obstacles and prevent collisions between UAVs in the formation during flight. The UAV swarm flight trajectory implemented by numerical simulation in MATLAB proves the effectiveness of this method.

[0009] This paper proposes a method for drone swarm formations that can maintain a preset formation and prevent collisions between drones in the formation. By introducing an artificial potential field method, it solves the problems of obstacle avoidance and potential collisions between drones in the formation during drone swarm flight.

[0010] This application proposes a multi-unmanned vehicle formation overall obstacle avoidance method based on an improved artificial potential field method. This method analyzes the force acting on the leader when there are obstacles on both sides. By introducing an additional repulsive force, the leader, under the influence of the combined force, moves to the center of the passable area and then moves forward from the center of the passable area until it is out of the obstacle influence area, ensuring that the two followers have the largest passable area. The formation is adjusted according to the width of the passable area. This scheme makes the present invention have the characteristics of stable formation and low switching frequency. This invention can effectively solve the problem of high formation switching frequency in obstacle environments.

[0011] Patent number: CN202210735335.2, patent name: Multi-robot formation obstacle avoidance control method based on improved artificial potential field method, which includes: establishing a multi-robot formation geometry structure in a virtual leader-leader-follower mode; when the multi-robot formation moves in a desired queue configuration, the multi-robot formation is controlled by the improved artificial potential field method and an obstacle avoidance control law is output; the output linear velocity and angular velocity of the formation robots are obtained according to the obstacle avoidance control law, and the formation robots are controlled to avoid obstacles while moving; in the improved artificial potential field method, the direction of the "repulsive force combined velocity" synthesized by the "repulsive force velocity" of the obstacle and the "repulsive force velocity" of the road boundary is adjusted to enable the formation robots to avoid obstacles within the safe area of ​​the road. The present invention effectively solves the problems of local optimality and unreachable targets existing in traditional artificial potential field methods.

[0012] It uses an improved artificial potential field method to perform obstacle avoidance control for multi-robot formations and output the obstacle avoidance control law, thereby obtaining the output linear velocity and angular velocity of the formation robots, and controlling the formation robots to avoid obstacles during operation; and by adjusting the direction of the "repulsive combined velocity" synthesized by the "repulsive velocity" of the obstacle and the "repulsive velocity" of the road boundary, it solves the problems of local optimal solution and unreachable target existing in the traditional artificial potential field method.

[0013] This application proposes a multi-unmanned vehicle formation obstacle avoidance method based on an improved artificial potential field method. This method analyzes the force acting on the leader when faced with obstacles on both sides. By introducing additional repulsive forces, the leader is forced to move to the center of the traversable area under the combined force, providing the largest traversable area for the two followers. The overall formation is adjusted based on the width of the traversable area, making the formation more stable and reducing the frequency of formation switching in multi-obstacle environments. This effectively solves the problem of high formation switching frequency in obstacle environments.

[0014] To address the frequent interruptions of formation missions caused by high obstacle avoidance frequency and formation changes in complex environments for multi-unmanned vehicle formations, this application proposes a comprehensive obstacle avoidance method for multi-unmanned vehicle formations in complex environments based on an improved artificial potential field method and a pilot-following method. First, a formation database that can be expressed using a unified formula is established. The accessible paths and widths of obstacle areas are determined using a potential field function. The resultant force applied to the pilot is adjusted based on the information obtained by the pilot, enabling reasonable and efficient obstacle avoidance. Once the last unmanned vehicle in the formation avoids the obstacle, the formation is restored and the mission continues. Summary of the Invention

[0015] To address these issues, a multi-AV platooning obstacle avoidance method based on an improved artificial potential field method is proposed. Combining the improved artificial potential field method with a pilot-follower approach, this method introduces an additional repulsive force to balance the horizontal force on the pilot, effectively utilizing terrain to achieve formation avoidance. This solves the problem of frequent formation changes in complex environments encountered by most other approaches.

[0016] To achieve the above object, the technical solution adopted by the present invention is:

[0017] A method for overall obstacle avoidance of multiple unmanned vehicle formations based on an improved artificial potential field method, comprising the following specific steps, characterized in that:

[0018] Step 1: Establish a commonly used formation database;

[0019] Step 2: Define attraction and repulsion, and obtain the functional relationship between the distance between the obstacle and the formation leader and repulsion. Calculate the distance between the leader unmanned vehicle and the obstacle, and then determine the formation width in the obstacle area that the formation can pass through.

[0020] Step 2 is as follows:

[0021] The distance between the navigator and the obstacle is obtained through the sensors carried by the unmanned vehicle:

[0022] The expression of gravity is:

[0023] F a (x) = -kd des

[0024] Where k is the gravitational gain coefficient, d des Indicates the distance between the pilot and the target point.

[0025] The expression formula of repulsive force:

[0026]

[0027] Where η is the repulsive force scale factor, d obs Represents the distance between the navigator and the obstacle, ρ0 represents the influence radius of each obstacle; at the same time, the navigator obtains d obs1 with d obs2 The distance d between the two obstacles is calculated based on the size of θ1 and θ2. The formula is as follows:

[0028]

[0029] Among them, d obs1 with d obs2 are the distances between the navigator and the obstacle, θ1 and θ2 are the angles between the line connecting the navigator and the target point and the obstacle, respectively;

[0030] Step 3: Define the additional repulsive force and obtain the functional relationship between the distance between the obstacle and the formation leader and the additional repulsive force;

[0031] Step 3 is as follows:

[0032] The expression formula of additional repulsion is:

[0033]

[0034] Among them, flag is the obstacle situation in front of the pilot car. When there are obstacles in front of the left and right sides of the pilot car, flag = 1; when obstacles are detected on only one side, flag = 0; F a (x) is the gravitational force of the target point, θ is the complementary angle between the gravitational force of the target point and the angle bisector of the angle formed by the two obstacles, and the trigonometric function relationship is derived as follows:

[0035] The net force on the navigator under the influence of gravity, repulsion, and additional repulsion is:

[0036] F(x)=F a (x)-(F re1 (x)+F re2 (x)+F r (x))

[0037] Among them, F re1 (x) is the repulsive force of obstacle 1 on the navigator, F re2 (x) is the repulsive force of obstacle 2 on the navigator;

[0038] When the navigator moves to the middle of the two obstacles under the resultant force, the horizontal component of the repulsive force is offset, and the horizontal component of the attractive force is offset by the additional repulsive force. The navigator is only subjected to the vertical force, and the magnitude of the resultant force is:

[0039]

[0040] Step 4: Obtain the formation width for the formation to pass through the obstacle area obtained in step 2, compare it with the current formation, add the repulsive force in step 2 and the additional repulsive force in step 3 to the navigator based on the position of the navigator, and select a reasonable formation from the formation database established in step 1 to switch to ensure smooth passage of the obstacle;

[0041] Step 5: After the last unmanned vehicle in the formation passes through the obstacle area, the initial formation is restored.

[0042] As a further improvement of the present invention, step 1 is specifically as follows:

[0043] A formation database is established for commonly used formations in formation control, and the pilot-follower method is used for formation control. In order to represent the relationship between unmanned vehicles and formation parameters, the general formula for the parameter information matrix that defines the formation shape is as follows:

[0044] F t =[F x1 F x2 ... F xn ] 4×n

[0045] F xj =[f 1j f 2j f 3j f 4j ] T

[0046] j=1,…,n,f 1j =j,f 2j =i,

[0047] Among them F t Represents the specific parameter information matrix of the t-th formation, F xj Represents the jth unmanned vehicle U j Formation information: F xj It consists of 4 parts: 1j The unmanned vehicle U in the specific formation j Number; f 2j Indicates U j The number of the leader being followed in the formation; f 3j Indicates following the unmanned vehicle U j the desired distance to be maintained between it and its pilot; Represented as unmanned vehicle U i with U j The desired distance to be maintained between 4j To follow the unmanned vehicle U j The desired heading angle to be maintained with its pilot; For unmanned vehicles i with U j The desired direction angle to maintain between them.

[0048] As a further improvement of the present invention, step 4 is specifically as follows:

[0049] The obstacle avoidance strategy is selected based on the size of d and d0, and the obstacle avoidance formation is selected from the formation database. When d0<d<2d0, the formation needs to choose a straight-line formation to pass the obstacle; when 2d0<d<d1, the formation needs to choose to reduce the original formation to pass the obstacle; when d>d1, the formation can maintain the original formation to pass the obstacle; where d1 is the initial formation width and d0 is the leader width.

[0050] As a further improvement of the present invention, step 5 is specifically as follows: when the last unmanned vehicle in the formation passes through an obstacle, the formation changes to the initial formation through the pilot-follow method, and the navigator restores the formation through the pilot-follow method based on the formation database and continues to perform the mission.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention forms a formation through a pilot-following method, adopts a navigator to control the formation, establishes a commonly used formation database, uses the navigator's obstacle distance to obtain the passable width of the obstacle area, introduces an additional repulsive force to keep the navigator in the middle of the passable area, formulates a reasonable formation obstacle avoidance strategy based on the comparison result of the passable area width and the formation width, reduces the formation transformation, and restores the formation through the pilot-following method after the obstacle avoidance is completed, thereby reducing the problem of low formation task completion rate in complex environments caused by each unmanned vehicle avoiding obstacles on its own and deviating from the expected trajectory in other solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a flow chart of the formation obstacle avoidance method;

[0054] Figure 2 Schematic diagram of the formation obstacle avoidance strategy. DETAILED DESCRIPTION

[0055] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and examples.

[0056] like Figure 1 As shown, a method for overall obstacle avoidance of multiple unmanned vehicle formations in complex environments based on an improved artificial potential field method and a pilot-following method includes the following steps:

[0057] (1) A database is established for the formations commonly used in formation control. The pilot method is used for formation control. In order to represent the relationship between the unmanned vehicles and the formation parameters, the general formula for the parameter information matrix that defines the formation shape is as follows:

[0058] F t =[F x1 F x2 … F xn ] 4×n

[0059] F xj =[f 1j f 2j f 3j f 4j ] T

[0060] j=1,…,n,f 1j =j,f 2j =i,

[0061] Among them F t Represents the specific parameter information matrix of the t-th formation, F xj represents the U of the j-th autonomous vehicle j Formation information: F xj It consists of 4 parts: 1j The unmanned vehicle U in the specific formation j Number; f 2j Indicates U j The number of the leader being followed in the formation; f 3j Indicates follower U j the desired distance to be maintained between it and its pilot; Represented as unmanned vehicle U i with U j The desired distance to be maintained between 4j For followers U j The desired heading angle to be maintained with its pilot; For unmanned vehicles i with U j The desired direction angle to maintain between them.

[0062] (2) When the leader in the formation detects an obstacle, it enters the obstacle avoidance mode: the leader determines the number of obstacles and the distance between obstacles based on the potential function, and matches the appropriate formation from the formation database established in (1).

[0063] like Figure 2 As shown, d obs1 with d obs2 The size of can be obtained through the sensors of the unmanned vehicle, and the width between obstacles for the unmanned vehicle to pass can be obtained through the distance between each obstacle and the navigator.

[0064] The expression of gravity is:

[0065] F a (x) = -kd des

[0066] Where k is the gravitational gain coefficient, d des Indicates the distance between the pilot and the target point.

[0067] The expression formula of repulsive force:

[0068]

[0069] Where η is the repulsive force scale factor, dobs represents the distance between the object and the obstacle, and ρ0 represents the influence radius of each obstacle;

[0070] At the same time, the navigator obtains d obs1 with d obs2 The distance d between the two obstacles is calculated based on the size of θ1 and θ2. The formula is as follows:

[0071]

[0072] Among them, F1 and F2 are the repulsive forces of obstacles on the unmanned vehicle in the potential field, d obs1 with d obs2 are the distances between the navigator and the obstacle, θ1 and θ2 are the angles between the line connecting the navigator and the target point and the line connecting the navigator and the obstacle, respectively.

[0073] (3) When the leader in the formation detects an obstacle, it enters obstacle avoidance mode: the leader adjusts its position according to the gravitational force of the target object.

[0074] like Figure 2 As shown in Figure 2, the functional relationship of the navigator's additional repulsive force can be obtained through the target's gravitational force on the navigator.

[0075] The expression formula of additional repulsion is:

[0076]

[0077] Among them, flag is the obstacle situation in front of the pilot car. When there are obstacles in front of the left and right sides of the pilot car, flag = 1; when obstacles are detected on only one side, flag = 0; F a (x) is the gravity of the target point, θ is the angle between the gravity of the target point and the horizontal direction, and the trigonometric function relationship can be deduced

[0078] The net force on the navigator under the influence of gravity, repulsion, and additional repulsion is:

[0079] F(x)=F a (x)-(F re1 (x)+F re2 (x)+F r (x))

[0080] Among them, F re1 (x) is the repulsive force of obstacle 1 on the navigator, F re2 (x) is the repulsive force of obstacle 2 on the navigator.

[0081] When the navigator moves to the middle of the two obstacles under the resultant force, the horizontal component of the repulsive force is offset, and the horizontal component of the attractive force is offset by the additional repulsive force. The navigator is only subjected to the vertical force, and the magnitude of the resultant force is:

[0082]

[0083] The navigator car runs to the middle of the two obstacles under the action of the target point's gravity, the obstacle's repulsion, and the additional repulsive force. At this time, the horizontal repulsive forces of the two obstacles cancel each other out, and the horizontal component of the target point's gravity is offset by the additional repulsive force. The navigator car is only subjected to the vertical force and runs out of the obstacle area under the action of the vertical force.

[0084] (4) Select the obstacle avoidance strategy based on the size of d and d0, and select the obstacle avoidance formation from the formation database. When d0<d<2d0, the formation needs to choose a straight formation to pass the obstacle; when 2d0<d<d1, the formation needs to choose to reduce the original formation to pass the obstacle; when d>d1, the formation can maintain the original formation to pass the obstacle; where d1 is the initial formation width, and d0 is the width of the unmanned vehicle.

[0085] (5) When the last unmanned vehicle in the formation passes the obstacle, the formation needs to be changed from the obstacle avoidance formation to the initial formation. The navigator uses the formation database in step (1) to restore the formation through the pilot-following method and continue to perform the mission.

[0086] The above description is merely a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent variation based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for platooning multi-unmanned vehicles to avoid obstacles based on an improved artificial potential field method, comprising the following steps, characterized in that: Step 1: Establish a formation database; Step 2: Define attraction and repulsion, and obtain the functional relationship between the distance between the obstacle and the formation leader and repulsion. Calculate the distance between the leader unmanned vehicle and the obstacle, and then determine the formation width in the obstacle area that the formation can pass through. Step 2 is as follows: The distance between the navigator and the obstacle is obtained through the sensors carried by the unmanned vehicle: The expression of gravity is: ； in, is the gravitational gain coefficient, Indicates the distance between the navigator and the target point; The expression formula of repulsive force: ； in, is the repulsive force scale factor, Indicates the distance between the pilot and the obstacle, Represents the influence radius of each obstacle; at the same time, the navigator obtains and Size and and , find the distance between two obstacles , the formula is as follows: ； in, and are the distances between the pilot and the obstacle, and are the angles between the line connecting the navigator and the target point and the obstacle; Step 3: Define the additional repulsive force and obtain the functional relationship between the distance between the obstacle and the formation leader and the additional repulsive force; Step 3 is as follows: The expression formula of additional repulsion is: ； in, For the obstacle situation in front of the pilot car, when there are obstacles in front of the left and right of the pilot car, ; When an obstacle is detected on only one side, ; is the gravitational force of the target point, It is the complementary angle of the angle between the target point gravity and the angle bisector of the angle between the two obstacles, and is derived from the trigonometric function relationship. ; The net force on the navigator under the influence of gravity, repulsion, and additional repulsion is: ； in, is the repulsive force of obstacle 1 on the navigator, is the repulsive force of obstacle 2 on the navigator; When the navigator moves to the middle of the two obstacles under the resultant force, the horizontal component of the repulsive force is offset, and the horizontal component of the attractive force is offset by the additional repulsive force. The navigator is only subjected to the vertical force, and the magnitude of the resultant force is: ； Step 4: Obtain the formation width for the formation to pass through the obstacle area obtained in step 2, compare it with the current formation, add the repulsive force in step 2 and the additional repulsive force in step 3 to the navigator based on the position of the navigator, and select a reasonable formation from the formation database established in step 1 to switch to ensure smooth passage of the obstacle; Step 5: After the last unmanned vehicle in the formation passes through the obstacle area, the initial formation is restored.

2. The method for overall obstacle avoidance of a multi-unmanned vehicle formation based on an improved artificial potential field method according to claim 1, characterized in that: Step 1 is as follows: A formation database is established for formation control, and the pilot-follower method is used for formation control. In order to represent the relationship between unmanned vehicles and formation parameters, the general formula for the parameter information matrix that defines the formation shape is as follows: ； ； ； in Represents the specific parameter information matrix of the t-th formation, represents the jth unmanned vehicle Formation information: It consists of 4 parts: For unmanned vehicles in a specific formation Number; express the number of the leader being followed in the formation; Indicates following the unmanned vehicle the desired distance to be maintained between it and its pilot; Represents an unmanned vehicle and The desired distance to be maintained: To follow the unmanned vehicle The desired heading angle to be maintained with its pilot; For driverless cars and The desired direction angle to maintain between them.

3. The method for overall obstacle avoidance of a multi-unmanned vehicle formation based on an improved artificial potential field method according to claim 2, characterized in that: Step 4 is as follows: according to and Size selection obstacle avoidance strategy, select obstacle avoidance formation from the formation database, when When the formation needs to choose a straight line formation to pass the obstacle; when When , the formation needs to choose to reduce the original formation to pass the obstacle; when When the formation passes through the obstacle, it can keep the original formation; is the initial formation width, The width of the navigator.

4. The method for overall obstacle avoidance of a multi-unmanned vehicle formation based on an improved artificial potential field method according to claim 3, characterized in that: Step 5 is as follows: When the last unmanned vehicle in the formation passes the obstacle, the formation switches to the initial formation through the pilot-follow method. The navigator restores the formation through the pilot-follow method based on the formation database and continues to perform the mission.

Images