An Unmanned Vehicle Formation Control Method Based on a Consistent Joint Virtual Potential Field

Abstract

The present invention discloses a unmanned vehicle formation control method of a consistent combined virtual potential field. The method improves the artificial potential field method, and takes into account the inertia factor in the control stage, and also takes into account the consistency requirements of the distributed system. First, a two-dimensional dynamic model is built for the unmanned vehicles in the team, and the unmanned vehicle formation and the ground station adopt distributed communication. In response to the consistency requirements of the distributed system, each unmanned vehicle node needs to maintain consistency in position and speed; then, the easy-to-understand characteristics of the artificial potential field method are used to apply it to the unmanned workshop formation; finally, the total input of the unmanned vehicle is determined, which is composed of the formation control input and the target point movement input, and the ROS distributed communication mechanism is used to carry out information within the formation. The design of the target point movement input of the present invention fully reduces the influence of the inertia factor, thereby reducing the deviation of the physical movement test; the consistent formation method is combined with the virtual potential field to realize the functions of unmanned vehicle formation movement, formation change and formation maintenance.

Description

Technical Field

[0001] The present invention relates to a formation control method for unmanned vehicles with a consistent combined virtual potential field, belonging to the technical field of unmanned vehicles. Background Art

[0002] In recent years, unmanned vehicle clusters have become a hot research topic as an important part of the intelligent and unmanned trend. As a highly mobile unmanned mover, unmanned aerial vehicles (UAVs) have the characteristics of being small, low-cost, easy to carry and project, and thus are widely used in modern warfare, people's livelihood and economic life. Clustered unmanned vehicles can complete tasks such as combat against enemies, logistics distribution and disaster relief support more efficiently and quickly. Multi-unmanned vehicle control algorithms mainly include path planning, formation movement, obstacle avoidance and collision avoidance, as well as formation transformation, reconstruction and target detection and search. Formation algorithms such as the classic Leader-Follower method and artificial potential field method. Among them, the artificial potential field method has attracted many researchers because of its simple principle and easy implementation. However, there are currently three prominent problems with unmanned vehicles: 1) the distance between the target and the obstacle is too close, resulting in the target being unreachable; 2) the gravitational force and the repulsive force are equal in magnitude and opposite in direction, making it easy to fall into local minima; 3) in a narrow track area, repulsive forces are exerted by all surrounding obstacles, and the unmanned vehicle may fall into oscillation. The primary goal of using the artificial potential field method for cluster formation control is to solve these three prominent problems to ensure the stability and safety of the cluster. In response to the various shortcomings of the artificial potential field method, some scholars have proposed applying an additional force to make the unmanned vehicle break away from the current minimum point, or reducing the influence of the obstacle repulsive force when near the target to make the unmanned vehicle slowly approach the target. However, these methods all have problems such as instability and their practicality needs to be verified. When applied to the actual environment, due to various interferences (wind speed, electromagnetic interference, environmental noise) and inertial factors in the real environment, the safety of the unmanned vehicle cannot be fully guaranteed, thus affecting the stability of the cluster. In addition, due to the high mobility of the cluster, and the cluster system is a non-ideal system with high order, nonlinearity, communication delay and interference in a conventional environment. The cluster system needs to maintain consistent stability, define the position and velocity coordination variables for formation maintenance, and at the same time, for a distributed system, it needs to satisfy information consistency. Therefore, consistency is one of the primary considerations when designing a distributed cluster control algorithm. Based on the research in these two directions, combining the artificial potential field and the consistency method, some scholars have proposed a leader-follower method based on the artificial potential field, which takes into account both the formation effect and the consistency requirements. However, this method does not consider inertial factors, has a low environmental adaptability, and does not consider the influence of the unmanned vehicle level and the environmental position on the virtual potential field force. However, the above research is only limited to a certain stage of cluster formation simulation, and none of them have conducted a complete process analysis of the entire process of the cluster formation movement task in the real environment. Summary of the Invention

[0003] The object of the present invention is to provide a formation control method for unmanned vehicles with a consistent combined virtual potential field in view of the deficiencies of the above-mentioned existing technologies. This method improves the artificial potential field method, takes into account the inertial factors in the environment during the control stage, and also considers the consistency requirements of the distributed system. First, a two-dimensional dynamic model of the unmanned vehicles within the team is established, and the unmanned vehicle formation and the ground station use distributed communication. In response to the consistency requirements of the distributed system, each unmanned vehicle node needs to maintain the consistency of position and speed. Then, taking advantage of the easy-to-understand characteristics of the artificial potential field method, it is applied to the formation of unmanned vehicles, and at the same time, it is improved to solve the problem of unreachable targets. Finally, the total input of the unmanned vehicle is determined, which is jointly composed of the formation control input and the target point movement input, and the ROS distributed communication mechanism is used to update the information within the formation in real time.

[0004] The present invention adopts the following technical solutions to achieve the above-mentioned invention object: A formation control method for unmanned vehicles with a consistent combined virtual potential field, the method comprising the following steps:

[0005] Step 1: Before the unmanned vehicle starts the test, check that each sensor is working properly, conduct a cluster communication test. After the test is completed, unlock, start and hover, and monitor whether the parameters of each unmanned vehicle are normal;

[0006] Step 2: The ground station sends a task instruction through the data link, and the unmanned vehicle cluster receives the instruction, records the initial GPS and calculates the offset L and speed v;

[0007] Step 3: Each unmanned vehicle publishes its own state including position, speed and quaternion, subscribes to the state information of neighboring nodes, forms the corresponding formation after obtaining the formation offset, calculates the formation maintenance input, generates a potential function and calculates the potential function input;

[0008] Step 4: Calculate the formation control input from the potential function input and the formation maintenance input Calculate the target point movement input from the real-time speed and position, and finally calculate the total input and input it into the flight control.

[0009] Advantageous effects:

[0010] 1. By separating the repulsive force exerted by the obstacle on the unmanned vehicle, the present invention makes the target point movement input change with the change of the distance from the target point, reducing the displacement error.

[0011] 2. By adding the consistency requirements of the distributed system to the formation control strategy, the present invention enables the cluster to maintain the consistency of position and speed, realizes the functions of formation maintenance and formation transformation, and ensures that the cluster can execute the movement task orderly. Description of the Drawings

[0012] Figure 1This is the flowchart of the method of the present invention. Specific implementation manners

[0013] The following will clearly and comprehensively describe the present invention with reference to the accompanying drawings of the present invention.

[0014] As Figure 1 shown, the present invention proposes a formation control method for a consistency joint virtual potential field, and the method includes the following steps:

[0015] Step 1: Before the unmanned vehicle starts the test, check whether each sensor works normally, conduct a cluster communication test. After the test is completed, unlock, start and hover, and monitor whether the parameters of each unmanned vehicle are normal;

[0016] Step 2: The ground station sends a task instruction through the data link, and the unmanned vehicle cluster receives the instruction, records the initial GPS and calculates the offset L and speed v;

[0017] Step 3: Each unmanned vehicle publishes its own state including position, speed and quaternion, subscribes to the state information of neighboring nodes, forms the corresponding formation after obtaining the formation offset, calculates the formation maintenance input, generates the potential function and calculates the potential function input;

[0018] Step 4: Calculate the formation control input from the potential function input and the formation maintenance input Calculate the target point movement input from the real-time speed and position, and finally calculate the total input and input it into the flight control.

[0019] Formation control input The formula is:

[0020]

[0021] Wherein, is the potential function input, is the formation maintenance input;

[0022] The formula for the formation maintenance input is:

[0023]

[0024]

[0025] Wherein, p i , v i , u i are the expected position, speed and input item of the unmanned vehicle i,; v j , p j are the expected speed and expected position of the unmanned vehicle j, K ij is the formation coefficient between the unmanned vehicle i and the unmanned vehicle j, K p is the speed control coefficient, o iRepresents the formation offset of the driverless vehicle i, o j Represents the formation offset of the driverless vehicle j, and N is the number of driverless vehicles.

[0026] The formula for the input of the potential function of the driverless vehicle i is:

[0027]

[0028] Among them, U ij Is the potential energy between the driverless vehicle i and the driverless vehicle j, and the gravitational / repulsive force exerted by the potential field on the driverless vehicle is the gradient of the potential function; Represents taking the gradient of the position where the driverless vehicle i is located.

[0029] The formula for the potential field force exerted on the driverless vehicle i by the driverless vehicle j is:

[0030]

[0031] N represents the number of driverless vehicles, and f j Represents the potential field force exerted on the driverless vehicle i by the driverless vehicle j.

[0032] The virtual potential field function U ij (||p ij ||) The formula is:

[0033]

[0034] Among them, α, β, and δ are all constants, and p ij Is the relative distance between the driverless vehicle i and the driverless vehicle j; R represents the effective communication range of the driverless vehicle. The stability and robustness of this formula are deduced from existing literature. When the distance between driverless vehicles is too small, a repulsive force is applied to avoid collisions, and conversely, a certain gravitational force is applied to prevent driverless vehicles from leaving the group;

[0035] The formula for the input of the moving target point is:

[0036]

[0037]

[0038] Among them, s is the offset from the target point, which is a vector, and v i (s) represents the velocity vector of the driverless vehicle i at s; the meaning of zoom is the time attribute, and its value is the quotient of the maximum speed and the acceleration; v linit Represents the maximum speed; the input is jointly determined by the travel distance and speed attributes of the driverless vehicle to reduce the influence of inertia.

[0039] The total input u i The formula is:

[0040]

[0041] Among them, is the formation control input, is the target point movement input;

[0042] The first formation control input is composed of the superposition of the potential field function input of neighbor nodes and the formation keeping input, which ensures the consistency of the positions and speeds of the cluster and collision avoidance.

[0043] The present invention adopts a formation control strategy based on an improved artificial potential field method to maintain the stability of the distance between vehicles during formation movement. According to the consistency requirements of the distributed system, a formation keeping input is designed so that the formation can stably maintain the moving formation and can perform formation switching. At the same time, the target point input will change over time as the offset from the target point decreases, so as to reduce the influence of inertial factors in the real environment and reduce the displacement error. Specifically, it includes the following:

[0044] 1) The present invention improves the artificial potential field method, analyzes the reasons for the disadvantage of the unreachable target, strips the repulsive force exerted by the obstacle from the method, solves the problem of the unreachable target caused by the same repulsive force of the obstacle and the gravitational force of the target, and at the same time retains the internal potential field force of the formation. When the distance p ij between the unmanned vehicle and other vehicles in the team is less than the communication distance R, a repulsive force is applied to the unmanned vehicle. When the distance p ij between the unmanned vehicles is greater than R, an attractive force is applied to the unmanned vehicle to maintain the formation inside the cluster. Usually, after the unmanned vehicle reaches the target point, a displacement error will occur due to inertia. On this basis, the present invention considers the influence of inertial factors in the real environment and introduces a zoom variable, so that during the movement of the unmanned vehicle, its target point movement input changes with the magnitude of the offset from the target, thereby reducing the experimental error caused by inertial factors.

[0045] 2) The present invention designs a control method that meets the consistency requirements of the distributed system. The distributed cluster control system requires the nodes in the cluster to maintain the consistency of speed and position. According to the consistency rule: the formation input received by the unmanned vehicle i is: K p is the speed control coefficient. Assuming that the relative speed between the current unmanned vehicle i and the unmanned vehicle j is too large or too small, the input is adjusted to ensure speed consistency while preventing the distance between the unmanned vehicles from being too close; The expression is K ij [(p i -p j )-o i -o j , which ensures the position consistency between the unmanned vehicle i and the unmanned vehicle j. K ij is the formation coefficient, which jointly controls the formation transformation with the formation offset o i .

[0046] 3) The present invention combines the virtual potential field method with the consensus control method, and at the same time takes inertia into consideration in the target point movement input. The total input u of the unmanned vehicle i i consists of the formation control input and the target point movement input jointly. Among them, the formation control input is composed of the superposition of the potential function input of the virtual potential field and the consensus formation maintenance input, ensuring the consistency of the cluster position and speed and collision avoidance; the target point movement input is jointly determined by the travel distance and speed attributes of the unmanned vehicle, reducing the influence brought by inertia. The meaning of zoom is the time attribute, and the value is the quotient of the maximum speed and the acceleration.

[0047] The above is only the preferred embodiment of the present invention. The present invention is not limited to the above embodiment. There may be local minor structural modifications and parameter adjustments during the implementation process. If various modifications or variations of the present invention do not depart from the spirit and scope of the present invention, and fall within the scope of the claims of the present invention and equivalent technical scope, then the present invention also intends to include these modifications and variations.

Claims

1. An unmanned vehicle formation control method based on a consistent joint virtual potential field, characterized in that, The method includes the following steps: Step 1: Before the unmanned vehicle starts up for pre-test inspection, each sensor works normally. Conduct cluster communication testing. After the testing is completed, unlock and start it, and monitor whether the parameters of each unmanned vehicle are normal; Step 2: The ground station sends task instructions through the data link. The unmanned vehicle cluster receives the instructions, records the initial GPS and calculates the offset L and speed v; Step 3: Each unmanned vehicle publishes its own state including position, speed and quaternion, subscribes to the state information of neighboring nodes, forms the corresponding formation after obtaining the formation offset, calculates the formation keeping input, generates a potential function and calculates the potential function input; Step 4: Calculate the formation control input from the potential function input and the formation maintenance input Calculate the target point movement input from the real-time speed and position, and finally calculate the total input, the formation control input The formula is as follows: Among them, is the input of the potential function, is the input for formation keeping; The formula for the formation keeping input is: where p i , v i , u i are the desired position, speed and input of the driverless vehicle i, v j , p j are the desired speed and desired position of the driverless vehicle j, K ij is the formation coefficient between the driverless vehicle i and the driverless vehicle j, K p is the speed control coefficient, o i represents the formation offset of the driverless vehicle i, o j represents the formation offset of the driverless vehicle j, and N is the number of driverless vehicles; The formula for the potential function input of unmanned vehicle i is: Among them, U ij is the potential energy between the driverless vehicle i and the driverless vehicle j, and the gravitational / repulsive force exerted by the potential field on the driverless vehicle is the gradient of the potential function; represents taking the gradient of the position where the driverless vehicle i is located; The formula for the potential field force exerted on unmanned vehicle i by unmanned vehicle j is: N represents the number of driverless vehicles, and f j represents the potential field force exerted by driverless vehicle j on driverless vehicle i; Virtual potential field function U ij (||p ij ||) The formula is as follows: where α, β, and δ are all constants, and p ij is the relative distance between autonomous vehicle i and autonomous vehicle j; R represents the effective communication range of the autonomous vehicle; The formula for the target point movement input is: where s is the offset from the target point, which is a vector, and v i (s) represents the velocity vector of the autonomous vehicle i at s; the meaning of zoom is the time attribute, and its value is the quotient of the maximum speed and the acceleration; v limit represents the maximum speed; the input is jointly determined by the travel distance and velocity attributes of the autonomous vehicle, reducing the influence brought by inertia.

2. The unmanned vehicle formation control method based on a consistent joint virtual potential field according to claim 1, characterized in that The total input u in step 4 i The formula is: Among them, is the formation control input, is the target point movement input; The first item of formation control input is composed of the superposition of the potential field function input of neighboring nodes and the formation keeping input, which ensures the consistency of the position and speed of the cluster and collision avoidance.

3. The unmanned vehicle formation control method based on a consistent joint virtual potential field according to claim 1, characterized in that, The method includes: First, establish a two-dimensional dynamic model for the unmanned vehicles within the team, and the unmanned vehicle formation and the ground station adopt distributed communication. To meet the consistency requirements of the distributed system, each unmanned vehicle node needs to maintain the consistency of position and speed. Then, utilize the easy-to-understand feature of the artificial potential field method and apply it to the formation of unmanned vehicles. At the same time, improve it to solve the problem of unreachable targets. Finally, determine the total input of the unmanned vehicle, which is jointly composed of the formation control input and the target point movement input, and use the ROS distributed communication mechanism to perform real-time sharing and updating of information within the formation.

Images