Omnidirectional robot group collaborative path planning navigation method and system and storage medium

Abstract

The invention discloses an omni-directional robot group collaborative path planning navigation method and system and a storage medium, and the method comprises the following steps: S1, collecting data received by a main robot from a sensor of the main robot, and constructing a global planning track; s2, the current actual distance between the master robot and the slave robot is calculated, and the slave robot and the master robot are controlled to keep a preset cooperative spacing range; s3, driving the master robot and the slave robot to move synchronously according to the global planning track; s4, collecting data received by the slave robot from a sensor of the slave robot, and locally optimizing a path planning track of the slave robot; and S5, repeatedly executing the steps S1-S4 until the main robot reaches the target point. According to the method, the path is planned by locally optimizing the path, the path of each slave robot is accurately controlled, and the slave robots can conveniently pass through narrow space channels in sequence without collision, so that the processing efficiency of multi-robot coordination control is improved.

Description

Technical Field

[0001] The present invention relates to the field of robot navigation and control technology, and in particular to a collaborative path planning and navigation method, system and storage medium for an omnidirectional robot group. Background Art

[0002] As artificial intelligence technology has entered its golden age of technological development, the development of mobile robots that integrate environmental perception, dynamic decision-making and planning has been pushed to new heights in many fields. Path planning is a process of inspecting target structures and environments using robots equipped with radar equipment, that is, calculating the robot's driving path. At present, researchers' research on robot path planning mainly focuses on single robots. In addition to single mobile robots, collaborative path planning of collaborative robots is also one of the current topics.

[0003] For example, the path planning method for multi-mobile robot collaboration based on the ROS system disclosed in application publication number CN117970928A, the multi-machine system adopted in the above scheme can decouple complex tasks, adapt to the solution of complex tasks, and at the same time make the system have task flexibility; it reduces the computing pressure of a single computer, thereby improving computing efficiency and having higher controllability, fault tolerance and robustness.

[0004] However, the above schemes are mainly applied in relatively wide spaces. For narrow space channels, it is difficult to accurately control the paths of each robot so that they do not collide. In addition, in traditional robot control systems, task allocation usually adopts centralized and distributed task allocation algorithms. There are many problems with these two algorithms. For example, the centralized task allocation algorithm requires a lot of computing resources to process task allocation. As the number of robots increases, the control difficulty increases. The distributed task allocation algorithm cannot globally optimize task allocation. Frequent communication between robots will cause network congestion or delays, affecting the timeliness of the system. Path planning usually adopts the A* algorithm based on graph search, but this algorithm needs to calculate the path all the time in a dynamic environment with obstacles, which increases the amount of calculation. Summary of the invention

[0005] The purpose of the present invention is to provide a collaborative path planning and navigation method, system and storage medium for an omnidirectional robot group to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The collaborative path planning and navigation method of an omnidirectional robot group comprises the following steps:

[0008] S1, collects the data received by the main robot from its own sensors and constructs a global planning trajectory;

[0009] S2. Calculate the current actual distance between the master robot and the slave robot. If the current actual distance is less than the preset collaborative distance range, drive the slave robot away from the master robot; if the current actual distance is greater than the preset collaborative distance range, drive the slave robot closer to the master robot.

[0010] S3, driving the master robot and the slave robot to move synchronously according to the global planning trajectory;

[0011] S4, collect the data received by the slave robot from its own sensor, check and calculate the real-time distance between the slave robot and the detected obstacle, if the real-time distance is less than the set safety distance, drive the slave robot to make a circular motion around the master robot with the master robot as the center, away from the obstacle; otherwise, continue to drive the slave robot to maintain the current planned trajectory;

[0012] S5. Repeat steps S1-S4 until the main robot reaches the target point.

[0013] As a further solution of the present invention, the master robot and the slave robot are both four-wheel omnidirectional mobile robots, the moving parts of the four-wheel omnidirectional mobile robots use Mecanum wheels, and the local speed of the four-wheel omnidirectional mobile robots in the x-axis direction in the local reference system is and the local velocity in the y-axis direction The relevant expression is:

[0014]

[0015] Among them, ω1, ω2, ω3 and ω4 are the angular velocities of each motor axis of the four-wheeled omnidirectional mobile robot, V1, V2, V3 and V4 are the speeds of each wheel of the four-wheeled omnidirectional mobile robot, L is the radius of the four-wheeled omnidirectional mobile robot, r 轮 is the wheel radius, W P is the angular velocity of the four-wheeled omnidirectional mobile robot rotating around its own center;

[0016] According to the rotation angle θ between the global reference frame (X, Y) and the local reference frame (x, y), the global speed of the four-wheeled omnidirectional mobile robot in the X-axis direction is Global speed in Y-axis direction Local velocity in the x-axis direction Local velocity in the y-axis direction The relevant expressions are:

[0017]

[0018]

[0019] As a further solution of the present invention, in step S4, the angular velocity ω of the slave robot in a circular motion around the master robot with the master robot as the center is expressed as:

[0020] V ax =V mx +ωD p sinδ

[0021] V ay =V my +ωD p cosδ;

[0022] Among them, V ax is the x-direction velocity vector required for angular motion of the robot, V ay is the y-direction velocity vector required for angular motion of the robot, V mx is the velocity vector of the main robot in the X direction, V my is the velocity vector of the master robot in the Y direction, δ is the angle between the master robot and the slave robot, and the preset coordination distance between the master robot and the slave robot is D p ;

[0023] The expression for the angular velocity direction of the slave robot making a circular motion around the master robot with the master robot as the center is:

[0024]

[0025] Among them, D t is the distance between the robot and the obstacle at the current moment, D t+1 is the distance between the robot and the obstacle at the next moment.

[0026] Omnidirectional robot group collaborative path planning and navigation system, including:

[0027] At least two numbered omnidirectional mobile robots are provided with a communication module and a sensor, wherein the sensor is used to detect obstacle data around the omnidirectional mobile robots; communication data is transmitted between the omnidirectional mobile robots through the communication module;

[0028] The omnidirectional mobile robot with the next number will take the omnidirectional mobile robot with the previous number as its master robot, and so on;

[0029] The omnidirectional mobile robot is further configured with a global path planning module and / or a local path planning module. The global path planning module constructs a global planning trajectory according to the data received by the first numbered omnidirectional mobile robot from its own sensor, and sends the global planning trajectory to the remaining omnidirectional mobile robots through the communication module. Each omnidirectional mobile robot performs collaborative movement according to the global planning trajectory and a preset collaborative spacing range.

[0030] The local path planning module checks and calculates the real-time distance between the slave robot and the detected obstacle based on the data received by the slave robot from its own sensor. If the real-time distance is less than the set safety distance, the local planning trajectory of the slave robot is constructed, and the local planning trajectory is sent to the omnidirectional mobile robot with the number behind. According to the local planning trajectory, the current slave robot moves in a circle around the master robot with the master robot as the center, away from the obstacle, and the omnidirectional mobile robot with the number behind performs collaborative movement according to the latest received local planning trajectory and the preset collaborative spacing range; otherwise, the slave robot continues to run on the current planning trajectory;

[0031] Each omnidirectional mobile robot executes the above-mentioned omnidirectional robot group collaborative path planning and navigation method.

[0032] A storage medium stores computer instructions, and when the computer instructions are executed by a processor, the above-mentioned collaborative path planning and navigation method is executed.

[0033] The collaborative control method of an omnidirectional robot group comprises the following steps:

[0034] F1. Verify whether the current task of the omnidirectional mobile robot has been defined. If not, execute the DQN task allocation algorithm module to allocate tasks. The first numbered omnidirectional mobile robot is used as the highest-level robot to allocate tasks to other omnidirectional mobile robots.

[0035] F2. After the current task of the omnidirectional mobile robot is defined, check whether the current task needs to be continued; if it needs to be executed, run the navigation DQN algorithm module, execute the aforementioned collaborative path planning and navigation method, and perform collaborative movement;

[0036] F3. In each step of executing the navigation DQN algorithm module, its effectiveness is continuously evaluated. If no collision occurs, the iterative process continues until each omnidirectional mobile robot completes the task and reaches the target point; if the omnidirectional mobile robots collide, the iterative task ends.

[0037] As a further solution of the present invention, the navigation reward mechanism of the navigation DQN algorithm module is designed as follows:

[0038]

[0039] Among them, D t-1 represents the distance between the obstacle and the robot at time t-1, D t Represents the distance between the obstacle and the robot at time t;

[0040]

[0041]

[0042] Among them, n robots represents the number of robots in the collaborative robot group, P M represents the position of the main robot, P iA Represents the position of the slave robot.

[0043] As a further solution of the present invention, the task allocation reward mechanism of the DQN task allocation algorithm module is designed as follows:

[0044]

[0045] Among them, d max ,d min and d sel They respectively represent the distance between the farthest optional robot and the highest-level robot, the distance between the nearest optional robot and the highest-level robot, and the distance between the currently selected robot and the highest-level robot.

[0046] in,

[0047]

[0048]

[0049] Among them, n robots Represents the number of robots in the collaborative robot group, D set Represents the distance between the mission target point and the main robot, D i-set Represents the distance between the i-th slave robot and the mission target point.

[0050] A storage medium stores computer instructions, and when the computer instructions are executed by a processor, the above-mentioned collaborative control method is executed.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows: in terms of multi-robot task allocation and path planning, the present invention can automatically optimize the task allocation strategy according to the robot's status and task priority in real time, avoiding the idle and waiting time caused by uneven task allocation in traditional methods, and at the same time can adapt to environmental changes and dynamically adjust task allocation; in a dynamic obstacle environment, the method and system of the present invention can re-plan the path in real time, reduce unnecessary motion trajectories of slave robots by locally optimizing the path planning trajectory, and can accurately control the path of each slave robot, so as to facilitate passing through narrow space channels in sequence without collision, thereby improving the processing efficiency of multi-robot coordinated control. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A flowchart of a collaborative path planning and navigation method for an omnidirectional robot group;

[0053] Figure 2 Schematic diagram of the geometric configuration of the four-wheel omnidirectional mobile robot of the present invention;

[0054] Figure 3 Schematic diagram of relative motion between the master robot and the slave robot in the present invention;

[0055] Figure 4 Schematic diagram of the coordinated movement of the master robot and the slave robot in the present invention;

[0056] Figure 5 A schematic diagram of the slave robot and the obstacle in the present invention;

[0057] Figure 6 A schematic diagram of the movement of the slave robot rotating clockwise around the master robot to avoid obstacles in the present invention;

[0058] Figure 7 A schematic diagram of the movement of the slave robot rotating counterclockwise around the master robot to avoid obstacles in the present invention;

[0059] Figure 8 It is a schematic flow chart of the collaborative control method of the omnidirectional robot group in the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] See also Figure 1 , a collaborative path planning and navigation method for an omnidirectional robot group includes the following steps:

[0062] S1, collects the data received by the main robot from its own sensors and constructs a global planning trajectory;

[0063] S2. Calculate the current actual distance between the master robot and the slave robot. If the current actual distance is less than the preset collaborative distance range, drive the slave robot away from the master robot; if the current actual distance is greater than the preset collaborative distance range, drive the slave robot closer to the master robot.

[0064] S3, driving the master robot and the slave robot to move synchronously according to the global planning trajectory;

[0065] S4, collect the data received by the slave robot from its own sensor, check and calculate the real-time distance between the slave robot and the detected obstacle, if the real-time distance is less than the set safety distance, drive the slave robot to make a circular motion around the master robot with the master robot as the center, away from the obstacle; otherwise, continue to drive the slave robot to maintain the current planned trajectory;

[0066] S5. Repeat steps S1-S4 until the master robot and the slave robot reach the target point.

[0067] The method of the present invention can re-plan the path of the slave robot in real time, reduce unnecessary motion trajectories of the slave robot by locally optimizing the path planning trajectory, and can accurately control the path of each slave robot, so as to facilitate passing through narrow space channels in sequence without collision, thereby improving the processing efficiency of multi-robot coordinated control.

[0068] In an embodiment of the present invention, the master robot and the slave robot are both four-wheel omnidirectional mobile robots. The moving parts of the four-wheel omnidirectional mobile robots use Mecanum wheels, which ensure that the omnidirectional robot group can form motion configurations of multiple angles and directions in a limited space. Compared with differential drive wheeled robots, it is easier to move and turn.

[0069] The geometric configuration of the four-wheel omnidirectional mobile robot in the present invention is as follows: Figure 2 As shown in the figure, the wheel angle of the four-wheel omnidirectional mobile robot is 90°, and each omnidirectional wheel and its motor shaft have the same rotation axis; the speed of each wheel is V1, V2, V3 and V4, and according to these speeds, the driving direction of the unit vector is defined as D1, D2, D3 and D4, respectively. For each wheel, the angular velocity of the motor shaft is ω1, ω2, ω3 and ω4, and counterclockwise rotation is regarded as positive rotation.

[0070] It is associated with a local reference system (x,y) whose origin is given by Figure 1 The center of mass of the omnidirectional robot is determined, the x-axis of the local reference system points from the center of mass of the omnidirectional robot to the center of the fourth wheel, and the y-axis of the local reference system points from the center of mass of the omnidirectional robot to the center of the third wheel. The rotation angle between the global reference system (X, Y) and the local reference system (x, y) is θ, the radius of the omnidirectional robot is set to L, and the X-axis position vector of the center of mass of the omnidirectional robot in the global reference system is defined as P ox The Y-axis position vector of the center of mass of the omnidirectional robot in the global reference system is defined as P oy , for the i-th wheel, in the global reference system, r i is designated as the position vector of the omnidirectional robot’s center of mass pointing to the i-th wheel. For a four-wheeled omnidirectional mobile robot platform, Defined as the local velocity value in the x-axis direction of the local reference system, is defined as the local velocity value in the y-axis direction of the local reference system, Defined as the global velocity value in the X-axis direction in the global reference system, V y w is defined as the global velocity value in the Y-axis direction in the global reference system.

[0071] The expression of the rotation transformation matrix R is:

[0072]

[0073] Among them, γ is the rotation transformation angle;

[0074] Using the rotation transformation matrix given by formula (1), we can obtain the position P of the i-th wheel relative to the center of mass in the local reference system: 0i , as shown below:

[0075]

[0076] From formula (2), we can get that the positions of each wheel relative to the center of mass in the local reference system are:

[0077]

[0078] The specific expression of the driving direction of the unit vector, that is, the driving direction of the i-th motor relative to the center of the omnidirectional robot in the local reference system is:

[0079]

[0080] Then the position vector r of the i-th wheel relative to the center of mass in the global reference system is i for:

[0081]

[0082] Derivation of formula (5) yields:

[0083]

[0084] Where x is the coordinate of the center of mass of the omnidirectional robot on the X axis in the global reference system, and y is the coordinate of the center of mass of the omnidirectional robot on the Y axis in the global reference system;

[0085] Therefore, the individual speed formula of each wheel of the four-wheeled omnidirectional mobile robot in the global reference system is obtained as follows:

[0086]

[0087] According to the local reference system, when θ = 0, the speed conversion matrix of the four wheels is calculated as shown in formula (8):

[0088]

[0089] Among them, W P is the angular velocity of the four-wheeled omnidirectional mobile robot rotating around its own center;

[0090] Based on formula (8), the inverse kinematics equation of the four-wheel omnidirectional mobile robot is:

[0091]

[0092]

[0093]

[0094]

[0095] According to the θ angle relationship, the global speed of the four-wheel omnidirectional mobile robot in the X-axis direction Global speed in Y-axis direction Local velocity in the x-axis direction Local velocity in the y-axis direction The relevant expressions are:

[0096]

[0097]

[0098] Among them, the linear speed conversion adopts V = ω·r 轮 , where r 轮 is the wheel radius, and the expression for the angular velocity that needs to be transmitted to each motor shaft is:

[0099]

[0100] Among them, ω1, ω2, ω3 and ω4 are the angular velocities of each motor axis of the four-wheeled omnidirectional mobile robot, V1, V2, V3 and V4 are the speeds of each wheel of the four-wheeled omnidirectional mobile robot, L is the radius of the four-wheeled omnidirectional mobile robot, r 轮 is the wheel radius, W P is the angular velocity of the four-wheeled omnidirectional mobile robot rotating around its own center;

[0101] In the present invention, the coordination distance between the master robot and the slave robot is preset to be D p , the collaborative distance error is β, then the preset collaborative distance range between the master robot and the slave robot is D p ±β, from the robot's motion state according to the coordination distance D p changes. Figure 3As shown in the figure, the master robot is fixed in the XY coordinate system and remains stationary. When the angular velocity is applied to the slave robot only around the center of mass of the master robot, the slave robot will rotate around the master robot with a coordinated spacing D. p The radius for angular motion.

[0102] The velocity vector required for the robot's angular motion is V0 = ω·D p , and the decomposition vector on the XY coordinate system is: V ax =V0 cosδ and V ay =V0 sinδ.

[0103] The position angle δ between the master robot and the slave robot is obtained by equation (16):

[0104]

[0105] Therefore, if Figure 4 As shown, when the master robot is in dynamic operation, the angular velocity ω of the slave robot in step S4 making a circular motion around the master robot with the master robot as the center is expressed as:

[0106] V ax =V mx +ωD p sinδ

[0107] V ay =V my +ωD p cosδ (17);

[0108] Among them, V ax is the velocity vector in the X direction required for the robot to move angularly, V ay is the Y-direction velocity vector required for angular motion of the robot, V mx is the velocity vector of the main robot in the X direction, V my is the velocity vector of the master robot in the Y direction, δ is the angle between the master robot and the slave robot;

[0109] The trajectory planning trajectory point array of the main robot from the starting point to the target point is shown in formulas (19) and (20):

[0110] p i ,i∈[0,1,2,...,n],p≡(p0,p1,p2,....p n ), p i =(x i ,y i ) (18);

[0111] Among them, p irepresents each point in the planned trajectory, and p represents the trajectory curve composed of each point in the planned trajectory;

[0112] path M ={(x0,y0),(x1,y1),...,(x i ,y i )} (19);

[0113] Among them, path M Represents the trajectory curve planned by the main robot;

[0114] The trajectory point array of the robot's trajectory planning is shown as follows:

[0115] path Aj ={(x0,y0),(x1,y1),...,(x i ,y i )},j∈[0,1,2,...,n] (20);

[0116] Among them, path Aj Represents the trajectory curve planned by the jth slave robot.

[0117] In the process of coordinated movement, Figure 5 As shown in the figure, the slave robot always follows the master robot at a coordinated distance. At the same time, in its local path planning module, the slave robot takes its own center of mass as the center and scans the possible obstacles with a scanning range of radius H. The position of the slave robot itself is (x si ,y si ), each trajectory point in its planned trajectory is (x i ,y i ), the path length between each trajectory point and the robot's center of mass is obtained by the following formula:

[0118]

[0119]

[0120] According to formula (22), each trajectory point with a path length greater than H is deleted, and each trajectory point with a path length less than or equal to H is retained. In this way, the local planning trajectory of the slave robot is limited to the sensor scanning range of the slave robot.

[0121] When the slave robot recognizes an obstacle through the laser sensor and calculates the real-time distance between the slave robot's own position and each obstacle according to formula (23), if the real-time distance is less than the set safety distance D0, the slave robot and the master robot maintain a coordinated distance D pUnder the premise of , the slave robot dynamically recalculates the angular velocity ω and the direction of movement, so as to control the slave robot to stay away from obstacles and avoid collision with them.

[0122]

[0123] Among them, (x Di ,y Di ) represents the coordinates of the obstacle surface point cloud data, (x si ,y si ) represents the real-time position coordinates of the slave robot;

[0124] In order to determine the angular velocity ω movement direction of the slave robot, it is necessary to determine the relative position of the obstacle. According to the relative position, the angular velocity direction applied to the slave robot is determined. The angular velocity direction of the slave robot in a circle around the master robot with the center of mass of the master robot as the center of the circle is expressed as follows:

[0125]

[0126] Among them, D t is the distance between the robot and the obstacle at the current moment, D t+1 is the distance between the robot and the obstacle at the next moment.

[0127] like Figure 6 As shown in the figure, during the coordinated motion of the slave robot and the master robot, the slave robot is located at the lower right of the master robot, and the obstacle is located in the motion trajectory above the slave robot. When the real-time distance between the slave robot's own position and each obstacle is less than the set safety distance D0, the slave robot and the master robot maintain a coordinated distance D p Under the premise of the above, the slave robot moves in a clockwise circle around the master robot to avoid obstacles.

[0128] like Figure 7 As shown in the figure, during the coordinated motion of the slave robot and the master robot, the slave robot is located at the lower left of the master robot, and the obstacle is located in the motion trajectory above the slave robot. When the real-time distance between the slave robot's own position and each obstacle is less than the set safety distance D0, the slave robot and the master robot maintain a coordinated distance D p Under the premise of the above, the slave robot moves in a counterclockwise circle around the master robot to avoid obstacles.

[0129] Omnidirectional robot group collaborative path planning and navigation system, including:

[0130] At least two numbered omnidirectional mobile robots are provided with a communication module and a sensor, wherein the sensor is used to detect obstacle data around the omnidirectional mobile robots; communication data is transmitted between the omnidirectional mobile robots through the communication module;

[0131] The omnidirectional mobile robot with the next number will take the omnidirectional mobile robot with the previous number as its master robot, and so on;

[0132] The omnidirectional mobile robot is further configured with a global path planning module and / or a local path planning module. The global path planning module constructs a global planning trajectory according to the data received by the first numbered omnidirectional mobile robot from its own sensor, and sends the global planning trajectory to the remaining omnidirectional mobile robots through the communication module. Each omnidirectional mobile robot performs collaborative movement according to the global planning trajectory and a preset collaborative spacing range.

[0133] The local path planning module checks and calculates the real-time distance between the slave robot and the detected obstacle based on the data received by the slave robot from its own sensor. If the real-time distance is less than the set safety distance, the local planning trajectory of the slave robot is constructed, and the local planning trajectory is sent to the omnidirectional mobile robot with the number behind. According to the local planning trajectory, the current slave robot moves in a circle around the master robot with the master robot as the center, away from the obstacle, and the omnidirectional mobile robot with the number behind performs collaborative movement according to the latest received local planning trajectory and the preset collaborative spacing range; otherwise, the slave robot continues to run on the current planning trajectory;

[0134] Each omnidirectional mobile robot executes the above-mentioned omnidirectional robot group collaborative path planning and navigation method to perform collaborative movement.

[0135] A storage medium stores computer instructions, and when the computer instructions are executed by a processor, the above-mentioned collaborative path planning and navigation method is executed.

[0136] For the collaborative control method of omnidirectional robot groups, please refer to Figure 8 , including the following steps:

[0137] F1. Verify whether the current task of the omnidirectional mobile robot has been defined. If not, execute the DQN task allocation algorithm module to allocate tasks; wherein the first numbered omnidirectional mobile robot is used as the highest-level robot to allocate tasks to other omnidirectional mobile robots;

[0138] F2. After the current task of the omnidirectional mobile robot is defined, check whether the current task needs to be continued to be executed to prevent other omnidirectional mobile robots in the group from being executed or completed; if there is a need to execute, run the navigation DQN algorithm module to execute the above-mentioned omnidirectional robot group collaborative path planning and navigation method to perform collaborative movement;

[0139] F3. At each step of the navigation DQN algorithm module, its effectiveness is continuously evaluated. If no collision occurs, the iterative process continues until each omnidirectional mobile robot completes the task and reaches the target point; if the omnidirectional mobile robots collide, the iterative task ends.

[0140] Distributed DRL (deep reinforcement learning) algorithms usually adopt adaptive reward mechanisms, leaderless environmental control methods and synchronous learning methods, and perform well in dealing with large-scale, high-complexity decision-making problems. On this basis, the present invention integrates the task allocation problem and the navigation problem into the DRL framework. In order to solve these two problems at the same time, the DQN (combining Q learning and deep neural network) algorithm module for navigation and task allocation is managed through the DRL algorithm framework, and the part that communicates with the agent and the environment is deleted. The algorithm runs inside each robot and interacts with the environment.

[0141] As a further solution of the present invention, the navigation reward mechanism of this embodiment focuses on four different situations: the master robot reaches the target, the slave robot with the number behind reaches the target, a collision occurs, and the robot pauses. The main purpose of navigation is to reach the target without a collision. Therefore, the navigation reward mechanism of the navigation DQN algorithm module is designed as follows:

[0142]

[0143] Among them, D t-1 represents the distance between the obstacle and the slave robot at time t-1, D t represents the distance between the obstacle and the slave robot at time t;

[0144] in,

[0145]

[0146]

[0147] Among them, n robots represents the number of robots in the collaborative robot group, P M represents the position of the main robot, P iA Represents the position of the slave robot.

[0148] In navigation tasks, the navigation reward mechanism can be designed based on the distance the robot is close to the target. The closer the distance to the target, the higher the reward. Reasonable reward design can help the robot reach the target point faster and safer, allowing the robot group to complete the task faster and better.

[0149] As a further solution of the present invention, the task allocation reward mechanism of this embodiment mainly includes two situations: selecting an idle target and selecting a priority target. The task allocation reward mechanism of the DQN task allocation algorithm module is designed as follows:

[0150]

[0151] Among them, d max ,d min and d sel They represent the distance between the farthest optional robot and the highest-level robot, the distance between the nearest optional robot and the highest-level robot, and the distance between the currently selected robot and the highest-level robot respectively;

[0152] in,

[0153]

[0154]

[0155] Among them, n robots Represents the number of robots in the collaborative robot group, D set Represents the distance between the mission target point and the main robot, D i-set Represents the distance between the i-th slave robot and the mission target point.

[0156] A reasonable task allocation reward mechanism can help the robot group clarify the task objectives. In the complex task allocation process, the reward mechanism can also be designed as a multi-objective optimization. It can be designed based on the distance of the robot to the target, and tasks can be assigned to the robot closest to the target first, thereby improving the efficiency of task completion and reducing unnecessary energy consumption of the robot group.

[0157] The method of the present invention takes the first-numbered omnidirectional mobile robot as the highest-level robot, which allocates tasks to other omnidirectional mobile robots through the DQ N task allocation algorithm module. Through the reasonable design of task allocation reward mechanism, the task allocation strategy can be automatically optimized in real time according to the robot status and task priority, avoiding the idle and waiting time caused by uneven task allocation in traditional methods, and at the same time can adapt to environmental changes and dynamically adjust task allocation.

[0158] By navigating the DQN algorithm module, we can not only ensure the safe operation of the robot by preventing collisions, but also quickly eliminate invalid paths through real-time collision feedback, speed up training, and ensure that the system can perform tasks more flexibly and reliably in a dynamic environment and maintain efficient collaboration.

[0159] A storage medium stores computer instructions, and when the computer instructions are executed by a processor, the above-mentioned collaborative control method is executed.

[0160] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A collaborative path planning and navigation method for an omnidirectional robot group, characterized in that: The steps include: S1, collects the data received by the main robot from its own sensors and constructs a global planning trajectory; S2. Calculate the current actual distance between the master robot and the slave robot. If the current actual distance is less than the preset collaborative distance range, drive the slave robot away from the master robot; if the current actual distance is greater than the preset collaborative distance range, drive the slave robot closer to the master robot. S3, driving the master robot and the slave robot to move synchronously according to the global planning trajectory; S4, collect the data received by the slave robot from its own sensor, check and calculate the real-time distance between the slave robot and the detected obstacle, if the real-time distance is less than the set safety distance, drive the slave robot to make a circular motion around the master robot with the master robot as the center, away from the obstacle; otherwise, continue to drive the slave robot to maintain the current planned trajectory; S5. Repeat steps S1-S4 until the main robot reaches the target point.

2. The omnidirectional robot group collaborative path planning and navigation method according to claim 1, characterized in that: The master robot and the slave robot are both four-wheel omnidirectional mobile robots. The moving parts of the four-wheel omnidirectional mobile robots use Mecanum wheels. The local speed of the four-wheel omnidirectional mobile robots in the x-axis direction in the local reference system is and the local velocity in the y-axis direction The relevant expression is: Among them, ω1, ω2, ω3 and ω4 are the angular velocities of each motor axis of the four-wheeled omnidirectional mobile robot, V1, V2, V3 and V4 are the speeds of each wheel of the four-wheeled omnidirectional mobile robot, L is the radius of the four-wheeled omnidirectional mobile robot, r 轮 is the wheel radius, W P is the angular velocity of the four-wheeled omnidirectional mobile robot rotating around its own center; According to the rotation angle θ between the global reference frame (X, Y) and the local reference frame (x, y), the global speed of the four-wheeled omnidirectional mobile robot in the X-axis direction is Global speed in Y-axis direction Local velocity in the x-axis direction Local velocity in the y-axis direction The relevant expressions are:

3. The omnidirectional robot group collaborative path planning and navigation method according to claim 1, characterized in that: In step S4, the angular velocity ω of the slave robot in a circular motion around the master robot with the master robot as the center is expressed as: Among them, V ax is the x-direction velocity vector required for angular motion of the robot, V ay is the y-direction velocity vector required for angular motion of the robot, V mx is the velocity vector of the main robot in the X direction, V my is the velocity vector of the master robot in the Y direction, δ is the angle between the master robot and the slave robot, and the preset coordination distance between the master robot and the slave robot is D p ; The expression for the angular velocity direction of the slave robot making a circular motion around the master robot with the master robot as the center is: Among them, D t is the distance between the robot and the obstacle at the current moment, D t+1 is the distance between the robot and the obstacle at the next moment.

4. Omnidirectional robot group collaborative path planning and navigation system, characterized in that: include: At least two numbered omnidirectional mobile robots are provided with a communication module and a sensor, wherein the sensor is used to detect obstacle data around the omnidirectional mobile robots; communication data is transmitted between the omnidirectional mobile robots through the communication module; The omnidirectional mobile robot with the next number will take the omnidirectional mobile robot with the previous number as its master robot, and so on; The omnidirectional mobile robot is further configured with a global path planning module and / or a local path planning module. The global path planning module constructs a global planning trajectory according to the data received by the first numbered omnidirectional mobile robot from its own sensor, and sends the global planning trajectory to the remaining omnidirectional mobile robots through the communication module. Each omnidirectional mobile robot performs collaborative movement according to the global planning trajectory and a preset collaborative spacing range. The local path planning module checks and calculates the real-time distance between the slave robot and the detected obstacle based on the data received by the slave robot from its own sensor. If the real-time distance is less than the set safety distance, the local planning trajectory of the slave robot is constructed, and the local planning trajectory is sent to the omnidirectional mobile robot with the number behind. According to the local planning trajectory, the current slave robot moves in a circle around the master robot with the master robot as the center, away from the obstacle, and the omnidirectional mobile robot with the number behind performs collaborative movement according to the latest received local planning trajectory and the preset collaborative spacing range; otherwise, the slave robot continues to run on the current planning trajectory; Each omnidirectional mobile robot executes the omnidirectional robot group collaborative path planning and navigation method described in any one of claims 1-3.

5. A storage medium, characterized in that: Computer instructions are stored, and when the computer instructions are executed by a processor, the collaborative path planning and navigation method as described in any one of claims 1-3 is executed.

6. A collaborative control method for an omnidirectional robot group, characterized in that: The steps include: F1. Verify whether the current task of the omnidirectional mobile robot has been defined. If not, execute the DQN task allocation algorithm module to allocate tasks. The first numbered omnidirectional mobile robot is used as the highest-level robot to allocate tasks to other omnidirectional mobile robots. F2. After the current task of the omnidirectional mobile robot is defined, check whether the current task needs to be continued; if it needs to be executed, run the navigation DQN algorithm module, execute the collaborative path planning and navigation method described in any one of claims 1 to 3, and perform collaborative movement; F3. In each step of executing the navigation DQN algorithm module, its effectiveness is continuously evaluated. If no collision occurs, the iterative process continues until each omnidirectional mobile robot completes the task and reaches the target point; if the omnidirectional mobile robots collide, the iterative task ends.

7. The omnidirectional robot group cooperative control method according to claim 6, characterized in that: The navigation reward mechanism of the navigation DQN algorithm module is designed as follows: Among them, D t-1 represents the distance between the obstacle and the robot at time t-1, D t Represents the distance between the obstacle and the robot at time t; Among them, n robots represents the number of robots in the collaborative robot group, P M represents the position of the main robot, P iA Represents the position of the slave robot.

8. The omnidirectional robot group collaborative control method according to claim 7, characterized in that: The task allocation reward mechanism of the DQN task allocation algorithm module is designed as follows: Among them, d max d min and d sel They represent the distance between the farthest optional robot and the highest-level robot, the distance between the nearest optional robot and the highest-level robot, and the distance between the currently selected robot and the highest-level robot respectively; in, Among them, n robots Represents the number of robots in the collaborative robot group, D set Represents the distance between the mission target point and the main robot, D i-set Represents the distance between the i-th slave robot and the mission target point.

9. A storage medium, characterized in that: Computer instructions are stored, and when the computer instructions are executed by the processor, the collaborative control method as described in any one of claims 6-8 is executed.

Images