Random walk robotic swarm control method, system, device, and medium

Abstract

This invention primarily relates to the field of intelligent robot technology. It discloses a method, system, device, and medium for controlling a robot swarm using random walks. The method includes: dividing the robot swarm into multiple robot groups, each group comprising multiple robots; determining a virtual leader within each robot group; for each robot group, each robot moving towards the virtual leader; each robot group performing a search operation using random walks; and when a robot group discovers a target object during its search, each robot within the group moving towards the target object to execute a capture task. This invention enables the exploration and capture of multiple targets while improving task efficiency.

Description

Technical Field

[0001] This invention relates to the field of intelligent robot technology, specifically to a method, system, device, and medium for controlling a random walk robot swarm. Background Technology

[0002] Multi-robot systems (MRS) consist of multiple autonomous robots that work collaboratively to complete tasks, exhibiting collective intelligence, scalability, and robustness that a single robot cannot achieve. They are widely used in exploration, rescue, and target containment applications. However, when controlling a swarm of robots to contain multiple targets using MRS, issues such as low search efficiency and long processing times arise. Therefore, further improvements are needed to enhance the ability of robot swarms to collaboratively solve problems in complex environments. Summary of the Invention

[0003] This invention provides a method, system, device, and medium for controlling a random-walking robot swarm, which can perform tasks of exploring and capturing multiple targets while improving the efficiency of task execution.

[0004] This invention provides a method for controlling a swarm of robots that perform random walks, the method comprising:

[0005] The robot swarm is divided into multiple robot groups, each of which includes multiple robots;

[0006] Determine a virtual leader for each of the robot groups;

[0007] For each of the robot groups, each robot moves with the virtual leader as its target;

[0008] Each of the robot groups conducted its search using a random walk approach.

[0009] When the robot group discovers a target object during a search operation, each robot in the robot group moves toward the target object to carry out a capture mission.

[0010] Furthermore, the provision that for each of the robot groups, each robot moves in response to the virtual leader includes:

[0011] Each robot moves in the direction of a preset composite potential field, wherein the composite potential field includes a first potential field and a second potential field.

[0012] When the robot detects the virtual leader, it generates an attraction based on the first potential field, causing the robot to move closer to the virtual leader under the influence of the attraction.

[0013] When the robot detects a target object, it generates a repulsive force based on the second potential field, causing the robot to move away from the target object under the action of the repulsive force, wherein the target object is the virtual leader, the robot, or an obstacle.

[0014] Furthermore, the random walk robot swarm control method further includes:

[0015] When the robot group is conducting a search operation in the current area, the density of the robot group in the current area is determined based on the time interval between the occurrence of interference events among the multiple robot groups. When two robot groups performing search operations interfere with each other, it is recorded as an interference event. The time interval is the duration between the occurrence times of the two interference events.

[0016] Furthermore, the random walk robot swarm control method further includes:

[0017] The total average time interval is determined based on the multiple time intervals mentioned above;

[0018] If the current time interval is greater than the total average time interval, then the step size of each robot in the robot group is reduced to narrow the search range of the robot group.

[0019] If the current time interval is less than the total average time interval, the step size of each robot in the robot group is increased to expand the search range of the robot group.

[0020] Furthermore, when the robot group discovers a target object during a search operation, each robot in the robot group moves with the target object as its target to perform a capture task targeting the target object, including:

[0021] Obtain the local location information of the robot group that discovered the target object, and assign each robot in the robot group that discovered the target object as a capture robot to perform a capture task against the target object;

[0022] Based on the local location information, a concentration field corresponding to the region where the target object is located is generated, wherein each location in the concentration field has a corresponding concentration value;

[0023] Based on a preset safety distance and the location information of the target object located within the concentration field, an equipotential line of concentration value is generated around the target object;

[0024] Each of the trapping robots is controlled to move closer to the target object, and the concentration value corresponding to the location of each trapping robot is obtained;

[0025] When the concentration value corresponding to the location of the trapping robot is equal to the concentration value of the equipotential line, the trapping robot is controlled to stop moving.

[0026] Furthermore, based on the local location information, a concentration field corresponding to the region where the target object is located is generated, wherein each location in the concentration field has a corresponding concentration value, including:

[0027] Based on the local location information, the area where the target object is located is divided into multiple grids;

[0028] The grid corresponding to the location of the target object is marked as the first grid, and the grid corresponding to the location of the obstacle is marked as the second grid;

[0029] Based on the location information and first distance of the first grid, a first concentration value is determined for each grid and a first concentration field is generated, wherein the first distance is the distance between the grids;

[0030] Based on the location information and the second distance of the second grid, a second concentration value is determined for each grid and a second concentration field is generated, wherein the second distance is the distance between the grids;

[0031] The first concentration field and the second concentration field are fused to obtain a fused third concentration field, and the third concentration value of each grid in the third concentration field is determined.

[0032] Furthermore, the aforementioned method for controlling a random walk robot swarm includes:

[0033] As the capture robot moves closer to the target object, the coordinate information of the capture robot is obtained;

[0034] Based on the coordinate information of the encirclement robot, a first offset of the encirclement robot is determined, wherein the first offset is the coordinate offset between the encirclement robot and another encirclement robot;

[0035] Based on the coordinate information of the target object and the coordinate information of the capture robot, a second offset of the capture robot is determined, wherein the second offset is the coordinate offset between the capture robot and the target object;

[0036] Based on the first offset and the second offset, a third offset of the capture robot is determined;

[0037] The movement of the capture robot is controlled based on the third offset.

[0038] The present invention provides a random walk robot swarm control system, the system including a control device and a robot swarm consisting of multiple robots;

[0039] A grouping module is used to divide a group of robots into multiple robot groups, each of which includes multiple robots;

[0040] The selection module is used to determine the virtual leader in each of the robot groups;

[0041] A first control module is configured to target each of the robot groups, with each robot moving toward the virtual leader.

[0042] The second control module is used to enable each robot group to perform search operations using a random walk approach;

[0043] The third control module is used to ensure that when the robot group discovers a target object during the search operation, each robot in the robot group moves towards the target object to perform a capture task for the target object.

[0044] Each of the robots is used to acquire distance information, coordinate information, and to identify target objects or obstacles.

[0045] The present invention also provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the random walk robot swarm control method as described in any of the preceding claims.

[0046] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the random walk robot swarm control method as described in any of the preceding claims.

[0047] The present invention has at least the following beneficial effects:

[0048] This application's technical solution improves task efficiency by dividing a robot swarm into multiple robot groups and assigning a virtual leader to each group, while simultaneously enabling the exploration and capture of multiple targets. Robots within each group move towards the virtual leader, maintaining group coordination and consistency. This structure enhances search efficiency because it allows multiple groups to search simultaneously in different areas, increasing the probability of target discovery. Furthermore, this solution employs a random walk approach for search operations. This strategy does not require prior environmental information, is suitable for unknown environments, and can cover a large area. Random walks provide flexibility, adapting to environmental changes and reducing repeated search areas, thus improving search efficiency. When any group discovers a target, all robots in that group quickly adjust their action strategies, concentrating their efforts to capture the target. This rapid response mechanism improves capture efficiency. Meanwhile, other groups continue their search tasks, ensuring that the overall search efficiency is not affected by the capture efforts of one group. Through this combination of distributed control and rapid response, efficient exploration and capture of multiple targets are achieved. Attached Figure Description

[0049] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the technical solutions of the present invention, and do not constitute a limitation on the technical solutions of the present invention.

[0050] Figure 1 This is a flowchart of the steps of the random walk robot swarm control method in this embodiment;

[0051] Figure 2 This is a flowchart of step S103 in the random walk robot swarm control method of this embodiment;

[0052] Figure 3 This is another step of the random walk robot swarm control method in this embodiment;

[0053] Figure 4 This is a flowchart of step S105 in the random walk robot swarm control method of this embodiment;

[0054] Figure 5 This is a flowchart of step S402 in the random walk robot swarm control method of this embodiment;

[0055] Figure 6 This is an algorithm architecture diagram of a robot swarm control method for random walks in an application scenario;

[0056] Figure 7 This is a schematic diagram illustrating the effect of a random walk robot swarm control method in an application scenario when performing an encirclement task;

[0057] Figure 8 This is a schematic diagram of a random walk robot swarm control system;

[0058] Figure 9 This is a flowchart illustrating the steps of the control device and the robot working together in the random walk robot swarm control system of this embodiment.

[0059] Figure 10 This is a schematic diagram of the structure of an electronic device. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0061] Before describing the technical solution of this application, the technical terms will be explained. For example:

[0062] Multi-Robot Systems (MRS) consist of multiple autonomous robots that work together to complete tasks, exhibiting collective intelligence, scalability, and robustness that a single robot cannot achieve. They are widely used in fields such as exploration, rescue, and target capture.

[0063] In the relevant technical field, the closest to the technical solution of this application is the exploration method proposed in "Bioinspired EnvironmentExploration Algorithm in Swarm Based on Lévy Flight and Improved Artificial Potential Field" [1], which combines the traditional Lévy flight random walk algorithm and the improved artificial potential field algorithm. This scheme guides the group of robots to explore the environment by using Lévy flight through a virtual leader, and uses the improved artificial potential field (APF) method to realize the formation maintenance, obstacle avoidance and flexible response to environmental changes among the group of robots. In this method, the movement of the virtual leader generates step size through the Lévy flight mechanism, guiding the group of robots to search gracefully and efficiently in the unknown environment, while ensuring that the robots maintain an appropriate distance to avoid collisions. The artificial potential field method is used to realize smooth obstacle avoidance behavior during the exploration process, so that the robot group can flexibly adapt to different terrains and obstacles like a natural group. A significant advantage of this method is that it does not rely on complex sensors and computing devices. The robot group can realize graceful search behavior through a simple random step size generation mechanism, just like natural organisms moving freely and changing formation in the environment. This method is particularly suitable for biomimetic robots that mimic the collective behavior of natural organisms.

[0064] However, the researchers in this application found that in the above-mentioned technical solutions, when the scale of the environment expands or the number of targets increases, a single team needs to spend a lot of time, which is inefficient and makes it impossible to carry out further encirclement tasks on multiple targets. For a multi-robot system, after finding a target, it should be able to further process the target. Although the Levi flight used in the above-mentioned technical solutions combines short-distance search with occasional large-step movements, which is suitable for the initial exploration of large areas, it is prone to frequent overlapping and physical interference in multi-robot tasks, resulting in low exploration efficiency. It is particularly inadequate in complex and dense environments and is not conducive to scaling up to multiple teams of robots to complete tasks collaboratively.

[0065] The reason for this is that the random nature of Levi's flight means that its step size and direction are largely unaffected by environmental factors. This means that even if the density of robots changes, its step size cannot be adjusted accordingly. Therefore, in environments with high robot density, maintaining a large step size leads to increased physical interference and path repetition, thus reducing overall exploration efficiency. In complex and frequently changing environments, the lack of this adaptive adjustment capability makes exploration strategies inflexible, easily leading to inefficient repetitive exploration and affecting the overall performance of the swarm of robots.

[0066] In related fields, even when attempting to introduce multiple robot squads, the lack of a systematic task allocation and cooperation mechanism means that each robot or squad still operates independently, employing an independent Lévy random exploration strategy. This approach fails to consider how to achieve a balanced task distribution among multiple squads, resulting in some areas being over-covered while others are missed. Furthermore, the lack of effective communication and cooperation makes it difficult for squads to form a unified encirclement operation after target discovery, failing to fully utilize the potential advantages of swarm robot systems in dynamic environments. Therefore, existing technological solutions fall short when facing multiple targets or tasks requiring complex cooperation.

[0067] In view of the fact that most traditional multi-robot systems focus on one independent aspect of target exploration, swarm movement, or target capture, and that the exploration methods are either limited by single-rule movement patterns and cannot effectively deal with dynamic targets, or limited by the limitations of traditional random walk algorithms and cannot balance exploration efficiency and cooperative operation, this application proposes the following embodiments to address the problem of efficient exploration and capture and rescue of multiple static or dynamic targets, and to perform the task of exploring and capturing multiple targets while improving the efficiency of task execution.

[0068] Please refer to Figure 1 , Figure 1 This is a flowchart of the steps of the random walk robot swarm control method in this embodiment.

[0069] This embodiment provides a method for controlling a random walk robot swarm, including:

[0070] S101. Divide the robot group into multiple robot subgroups, each of which includes multiple robots.

[0071] S102. Determine the virtual leader in each robot group.

[0072] S103. For each robot group, each robot moves with the virtual leader as its target.

[0073] S104. Each robot group conducts its search using a random walk method.

[0074] S105. When the robot group discovers a target object during a search operation, each robot in the robot group moves toward the target object to carry out a capture mission.

[0075] In step S101 of some embodiments, the robots are divided into groups based on their positions in the environment, for example, each group is responsible for a specific area or grid.

[0076] In step S102 of some embodiments, a leader is elected by an algorithm. For example, the robot with the strongest communication capability or the best position is selected as the leader, or the leader is rotated within the group to balance energy consumption and task allocation, or the leader is determined according to preset rules, such as ID number or startup order.

[0077] In step S103 of some embodiments, a following algorithm, such as a PID controller, is used to enable the robot to dynamically adjust its position to maintain its relative position with the leader. In addition, behavioral rules are set, such as maintaining a certain distance and avoiding collisions, to maintain the coordinated movement of the group. For example, the distance to the leader is monitored by sensors, and the speed and direction are automatically adjusted to maintain the formation.

[0078] It is understandable that this embodiment uses multiple teams for distributed collaborative exploration, which improves the system's efficiency in dealing with large environments and multiple objectives.

[0079] Please refer to Figure 2 , Figure 2 This is a flowchart of step S103 in the random walk robot swarm control method of this embodiment.

[0080] In some embodiments, step S103 includes:

[0081] S201. Each robot moves in the direction of a preset composite potential field, wherein the composite potential field includes a first potential field and a second potential field.

[0082] S202. When the robot detects the virtual leader, it generates gravity based on the first potential field, so that the robot moves closer to the virtual leader under the influence of gravity.

[0083] S203. When the robot detects a target object, it generates a repulsive force based on the second potential field to make the robot move away from the target object under the action of the repulsive force, wherein the target object is the virtual leader, the robot, or an obstacle.

[0084] Understandably, maintaining a safe distance between robots within the robot group, and between a robot and the virtual leader, involves repulsive forces. A robot approaches the virtual leader using gravitational pull, but upon reaching a certain point, it is constrained by repulsive forces to maintain a safe distance. Similarly, repulsive forces also exist between the robots within the group to maintain a safe distance.

[0085] In some embodiments, an artificial potential field algorithm is used to realize the swarm movement of a group of robots. In the artificial potential field method, the potential field in which the robot is located is artificially defined as a gravitational potential field (first potential field) and a repulsive potential field (second potential field). The gravitational potential field is provided by the target object, and the repulsive potential field is provided by the obstacle.

[0086] In this invention, an artificial potential field method is used to induce swarm movement among groups of robots. This gravitational field attracts individual robots within the group towards a virtual leader. The strength of the gravitational field is inversely proportional to the distance from the virtual leader; that is, the closer the robot is to the virtual leader, the weaker the gravitational force; the farther away, the stronger the gravitational force. The mathematical expression for the gravitational field is:

[0087]

[0088] in, Represents the gravitational field coefficient. Indicates the positions of other robots in the group. With the virtual leader of the group The distance between them This indicates the magnitude of the gravitational field.

[0089] The purpose of the repulsive field is to prevent collisions between team members while simultaneously avoiding obstacles. In other words, for other robots in the team, everything except the virtual leader is an obstacle. The strength of the repulsive field is inversely proportional to the distance from the obstacle; the closer to the obstacle, the stronger the repulsive force; the farther away, the weaker. The mathematical expression for the repulsive field is:

[0090]

[0091] in, Represents the repulsive field coefficient. Indicates the positions of other robots in the group. relative to the position of the obstacle The distance between them Indicates the maximum distance affected by the obstacle. This indicates the magnitude of the repulsive field.

[0092] The combined field, obtained by superimposing the effects of the gravitational and repulsive fields mentioned above, allows individual robots in each group to maintain a certain distance within the cluster and follow the virtual leader of the group when moving towards the direction of the combined field, while avoiding obstacles. To avoid getting trapped in local optima, an additional random perturbation term is added to the combined field. The mathematical expression for the combined field is:

[0093]

[0094] in, This represents a random disturbance term.

[0095] It is understood that in this embodiment, the squad is composed of robots with different functions. The leader robot uses the artificial potential field method to lead the members to achieve coordinated movement of the group, so as to maintain the consistency and flexibility of the squad and to adaptively avoid obstacles after they are discovered. At the same time, it provides a perturbation term for the artificial potential field method to avoid getting stuck in local minima during the obstacle avoidance process.

[0096] In some embodiments, each robot divides the space into several regions based on its sensing range and identifies the number of robots within each region. The robot density can be estimated by calculating the ratio of the number of robots within the sensing region to the total area. The advantage of this method is its ability to dynamically adjust the size of the sensing region to adapt to different environments and task requirements, thereby more accurately determining the density of the robot swarm.

[0097] In some embodiments, a specific implementation of the step of determining the density of robot groups in the current area includes:

[0098] When a robot group is conducting a search operation in the current area, the density of robot groups in the current area is determined based on the time interval between interference events that occur among multiple robot groups. When two robot groups performing search operations interfere with each other, it is recorded as an interference event, and the time interval is the duration between the occurrence times of the two interference events.

[0099] Please refer to Figure 3 , Figure 3 This is another step in the random walk robot swarm control method of this embodiment.

[0100] In some embodiments, the random walk robot swarm control method further includes:

[0101] S301. Determine the total average time interval based on multiple time intervals.

[0102] S302. If the current time interval is greater than the total average time interval, reduce the step size of each robot in the robot group when moving, so as to narrow the search range of the robot group.

[0103] S303. If the current time interval is less than the total average time interval, increase the step size of each robot in the robot group when moving to expand the search range of the robot group.

[0104] As we can understand, random walk refers to moving in random directions and with random step sizes. Random walk does not require prior knowledge of the environment map or the specific location of the target, making it well-suited for dynamic and unknown environments, where the target can also be moving randomly rather than remaining stationary or moving in a regular pattern.

[0105] Traditional random walk algorithms mainly include Brownian motion and Lévy flight. Brownian motion is suitable for local searches but not for efficient exploration of large areas; while Lévy flight combines short-distance searches with occasional large-step movements, making it suitable for initial exploration of large areas, it is prone to frequent overlapping and physical interference when extended to distributed movement of multiple leaders and squads, resulting in low exploration efficiency. Therefore, this embodiment adopts an improved random walk algorithm.

[0106] In the random walk algorithm used in this embodiment, each robot group can estimate its density based on the time interval between interference with other groups, i.e., estimate whether multiple robot groups are exploring the same local area. The step size is adaptively adjusted based on the density of other robot groups in the environment. When the robot group density is high, the step size decreases, searching the local area; when the density is low, the step size increases, searching a wider area. This method reduces redundant searches and improves search efficiency. The mathematical expression of this improved random walk algorithm is:

[0107]

[0108] in, Indicates the new step size. This indicates the previous step size. Indicates the robot's speed. This represents the total average time interval of physical interference between robot groups, used to estimate the density of robot groups. k represents an adjustment factor that controls the influence of the previous step size on the current step size. This indicates the time interval between the current interference event and the previous interference event.

[0109] when When the density of robot groups is low in a local area, the robots should increase their step length to cover a larger area; when When the density of robot groups is high in a local area, the robot should reduce its step size to avoid repeated searches with other groups.

[0110] Understandably, in an unknown environment, this embodiment improves exploration efficiency by adaptively adjusting the step size to reduce repetitive searches among robot groups.

[0111] In some embodiments, a random walk robot swarm control method further includes:

[0112] When the robot group is conducting a search operation in the current area, if the obstacle detected by the target robot in the robot group is a robot from the same robot group, the search operation is carried out according to the preset distance between the target robot and the robot from the same robot group. If the obstacle detected by the target robot is a robot from a different robot group, the target robot is controlled to move in a preset direction.

[0113] Understandably, to further ensure that each robot team ultimately converges to explore within its own independent area, when a robot from one team encounters another team and engages in obstacle avoidance, it turns in the opposite direction. In this way, the various groups using the artificial potential field method for swarm motion can evenly distribute themselves within the environment, enabling more effective environmental exploration, reducing redundant searches, and improving search efficiency.

[0114] Please refer to Figure 4 , Figure 4 This is a flowchart of step S105 in the random walk robot swarm control method of this embodiment.

[0115] In some embodiments, step S105 includes:

[0116] S401. Obtain the local location information of the robot group that discovered the target object, and assign each robot in the robot group that discovered the target object as a capture robot to perform the capture task against the target object.

[0117] S402. Based on local location information, generate a concentration field corresponding to the area where the target object is located, wherein each location in the concentration field has a corresponding concentration value.

[0118] S403. Based on the preset safety distance and the location information of the target object located within the concentration field, generate equipotential lines of concentration values ​​around the target object.

[0119] S404. Control each trapping robot to move closer to the target object and obtain the concentration value corresponding to the location of each trapping robot.

[0120] S405. When the concentration value corresponding to the location of the capture robot is equal to the concentration value of the equipotential line, control the capture robot to stop moving.

[0121] In this embodiment, after each group discovers the target, it switches from the original cluster movement exploration state to the encirclement state through a state machine, thereby removing the constraints of the artificial potential field method and the improved random walk method and adopting other control strategies to complete the encirclement task.

[0122] In some embodiments, a gene regulation network algorithm is employed to surround and capture the target after it has been found. The gene regulation network algorithm is a mechanism that simulates how genes control protein production and cell behavior in living organisms. In multi-robot systems, the swarm robot system uses local information to obtain the positions of targets and obstacles within the working area, thereby generating a concentration field for the targets and obstacles. Within this concentration field, the agent can calculate the concentration value at its current location based on its distance from the targets and obstacles. This process not only helps the agent move towards the target by descending the concentration gradient but also helps it avoid obstacles. An equipotential line is selected based on the minimum safe distance between the agent and the target. This equipotential line encircles the target (ignoring obstacles), acting as an encirclement. During the agent's movement towards the target, it stops moving once the concentration value at its location matches the concentration value of the encirclement.

[0123] Please refer to Figure 5 , Figure 5 This is a flowchart of step S402 in the random walk robot swarm control method of this embodiment.

[0124] In some embodiments, step S402 includes:

[0125] S501. Based on local location information, the area where the target object is located is divided into multiple grids.

[0126] S502. Mark the grid corresponding to the location of the target object as the first grid, and mark the grid corresponding to the location of the obstacle as the second grid.

[0127] S503. Based on the location information of the first grid and the first distance, determine the first concentration value of each grid and generate the first concentration field, wherein the first distance is the distance between the grids.

[0128] S504. Based on the location information of the second grid and the second distance, determine the second concentration value of each grid and generate a second concentration field, wherein the second distance is the distance between the grids.

[0129] S505. Merge the first concentration field and the second concentration field to obtain the merged third concentration field and determine the third concentration value of each grid in the third concentration field.

[0130] In this embodiment, the local environment detected by the team of robots that discovered the target is first meshed, with the grid containing the target and obstacles marked as "1" and the remaining grid marked as "0". Using the target location information, an initial concentration field for the target is generated. The specific formula is as follows:

[0131]

[0132]

[0133] Wherein, the distance from the i-th grid cell to the target is calculated as: . , These are the x and y coordinates of the i-th grid cell, respectively. , Let x and y be the x and y coordinates of the target, respectively. Calculate the effect of the target on the [missing information - likely a specific event or condition]. Concentration value of each grid Then, we can see that the concentration value is the highest at the target location, and the farther away from the target, the smaller the concentration value of the corresponding grid.

[0134] Similarly, using the obstacle location information, a raw concentration field for the obstacle is generated. The specific formula is as follows:

[0135]

[0136]

[0137] Wherein, the distance from the i-th grid cell to the obstacle is calculated as follows: . , These are the x and y coordinates of the i-th grid cell, respectively. , Let x and y be the coordinates of the obstacle, respectively. Calculate the concentration value of the i-th grid cell under the influence of the obstacle. .

[0138] The obtained raw concentration field is further processed to fuse the concentration fields formed by the target location information and the obstacle location information, ensuring that the concentration fields for the target and the obstacles are "opposite," thus guaranteeing that the selected equipotential lines (encircling loops) perfectly enclose the target. The specific formula is as follows:

[0139]

[0140]

[0141]

[0142]

[0143] in, It is a sigmoid function, which is commonly used in robot control systems to smoothly control the switching of certain behaviors. In the formula, x represents the current input value, and k and z represent the adjustment parameters.

[0144] When further processing the obtained original concentration field of obstacles The meaning is that at time t, after The concentration field formed by obstacles obtained after preliminary processing by the module , This includes the obstacle concentration values ​​corresponding to all grid cells, i.e., the original obstacle concentration field. 'k' represents the adjustment parameter.

[0145] When processing the obtained original concentration fields of the target and obstacles The meaning is that at time t, after The concentration field formed by the combined target and obstacle data obtained after initial processing by the module , It contains the target concentration values ​​corresponding to all grid cells, i.e., the original target concentration field. It contains the obstacle concentration values ​​corresponding to all grid cells, i.e., the original obstacle concentration field. 'k' represents the adjustment parameter.

[0146] The result and The final concentration field is obtained by fusion, and the equipotential line information of the encirclement mode used to surround the target is extracted (i.e., the concentration value corresponding to the equipotential line, which is determined by the set safe distance between the robot and the target). The meaning is that at time t, after The concentration field formed by the combined target and obstacle data obtained after initial processing by the module , 'k' represents the adjustment parameter.

[0147] In some embodiments, a method for controlling a random walk robot swarm further includes:

[0148] When the encirclement robot moves closer to the target object, its coordinate information is acquired. Based on the coordinate information of the encirclement robot, a first offset is determined, which is the coordinate offset between the encirclement robot and another encirclement robot. Based on the coordinate information of the target object and the coordinate information of the encirclement robot, a second offset is determined, which is the coordinate offset between the encirclement robot and the target object. Based on the first and second offsets, a third offset is determined. Based on the third offset, the movement of the encirclement robot is controlled.

[0149] It is understandable that the robot's coordinate system is set as a two-dimensional x-axis and y-axis coordinate system, but a three-dimensional coordinate system can also be set. In this embodiment, only a two-dimensional coordinate system is used as an example for illustration.

[0150] Specifically, the first offset is calculated using the following formula:

[0151]

[0152]

[0153] in, , Let x and y be the x and y coordinates of the i-th robot in the team, respectively. , These are the x and y coordinates of the j-th robot in the squad, respectively. , These represent the first offsets of the i-th robot on the x and y axes under the influence of the j-th robot.

[0154] Specifically, the second offset is calculated using the following formula:

[0155]

[0156]

[0157] in, , These are the x and y coordinates of the target, respectively. , These are the second offsets of the i-th robot on the x and y axes under the influence of the target, respectively.

[0158] Combining the effects of both factors, the robot's offset along the x and y axes can be obtained, as shown in the following formulas:

[0159]

[0160]

[0161] in, , These represent the offsets on the x and y axes, respectively, of the combined effects of the two actions on the robot target.

[0162] Understandably, the Gene Regulation Network (GRN) algorithm is used to construct a comprehensive concentration field between the team of robots, the target, and obstacles to complete the task of capturing the target, thus adapting to complex and ever-changing task environments.

[0163] Please refer to Figure 6 , Figure 6 This is an algorithm architecture diagram of a random walk robot swarm control method in an application scenario.

[0164] This embodiment aims to solve the problem of efficiently exploring and encircling multiple static or dynamic targets in complex environments where swarm robots are in unknown locations and where GPS communication is limited. For example... Figure 6 As shown, the proposed system mainly consists of three parts: by employing an artificial potential field algorithm, the virtual leader of the robot can lead other robots to form a small squad for cooperative movement and obstacle avoidance; by employing an improved random walk method that is superior to Levy flight, multiple robot squads can efficiently explore static or dynamic targets in the environment in a distributed manner; and by employing a gene regulatory network (GRN) algorithm to encircle and capture the discovered targets.

[0165] Please refer to Figure 7 , Figure 7 This is a schematic diagram illustrating the effect of a random walk robot swarm control method in an application scenario when performing an encirclement task.

[0166] like Figure 7 As shown, after the robot team discovers the target, all robots in the team switch from exploration mode to encirclement mode. The virtual leader in the team and the other robots with different functions under its leadership construct a comprehensive concentration field between the robot, the target, and obstacles using the Genetic Reactivity Network (GRN) algorithm to encircle the discovered target object. During the encirclement task, firstly, the robots in the team form an encirclement circle, surrounding the target object within the circle; then, based on the concentration value within the concentration field, each robot gradually moves closer to the target object, causing the encirclement circle to gradually shrink; finally, each robot stops moving at the equipotential line of the concentration value, thus completing the encirclement task.

[0167] Each of the above embodiments improves task efficiency by dividing the robot swarm into multiple robot groups and assigning a virtual leader to each group, while simultaneously enabling the exploration and capture of multiple targets. Robots within each group move towards the virtual leader, maintaining group coordination and consistency. This structure helps improve search efficiency because it allows multiple groups to search in different areas simultaneously, increasing the probability of target discovery. Furthermore, the search operation employs a random walk approach, a strategy that does not require prior environmental information, is suitable for unknown environments, and can cover a large area. Random walks make the search operation flexible, adaptable to environmental changes, and reduce repeated search areas, thereby improving search efficiency. When any group discovers a target, all robots in that group quickly adjust their action strategy, concentrating their efforts to capture the target. This rapid response mechanism enhances capture efficiency. Meanwhile, other groups continue their search tasks, ensuring that the overall search efficiency is not affected by the capture actions of one group. Through this combination of distributed control and rapid response, efficient exploration and capture of multiple targets are achieved.

[0168] Please refer to Figure 8 , Figure 8 This is a schematic diagram of a random walk robot swarm control system.

[0169] This embodiment also provides a random walk robot swarm control system including a control device 610 and a robot swarm consisting of multiple robots 620;

[0170] Grouping module 611 is used to divide the robot group into multiple robot subgroups, each of which includes multiple robots 620.

[0171] Select module 612 to determine the virtual leader in each robot group.

[0172] The first control module 613 is used for each robot group, where each robot moves with the virtual leader as its target.

[0173] The second control module 614 is used for each robot group to conduct search operations in a random walk manner.

[0174] The third control module 615 is used to ensure that when the robot group discovers a target object during a search operation, each robot in the robot group moves towards the target object to perform a capture task.

[0175] Each robot 620 is used to acquire distance information, coordinate information, and to identify target objects or obstacles.

[0176] Please refer to Figure 9 , Figure 9 This is a flowchart illustrating the steps of the control device and the robot working together in the random walk robot swarm control system of this embodiment.

[0177] For each robot controlled in a random walk robot swarm control system, firstly, the robot swarm to which each robot belongs initializes and constructs a unified local coordinate system within the swarm, and determines the swarm's overall initial orientation based on the task objective and environmental information. Secondly, within the swarm, movement is coordinated using an artificial potential field method. When exploring the environment using an improved stochastic method, collisions with other robots are detected and avoided through the robot's built-in sensors, and the step size is adjusted accordingly based on the robot density within the exploration area. Then, the presence of targets is continuously monitored and detected. Once a target is detected, a gene regulation network algorithm is used to participate in the formation of a containment pattern until the task is completed and operation ceases.

[0178] It will be understood by those skilled in the art that all or some of the steps and apparatuses in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. As is known to those skilled in the art, communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0179] It is understood that the content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0180] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement any of the above-mentioned random walk robot swarm control methods.

[0181] refer to Figure 10 , Figure 10 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0182] The processor 701 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0183] The memory 702 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 702 can store operating devices and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 702 and is called by the processor 701 to execute the random walk robot swarm control method of the embodiments of this application.

[0184] The input / output interface 703 is used to implement information input and output;

[0185] The communication interface 704 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0186] Bus 705 transmits information between various components of the device (e.g., processor 701, memory 702, input / output interface 703, and communication interface 704);

[0187] The processor 701, memory 702, input / output interface 703, and communication interface 704 are connected to each other within the device via bus 705.

[0188] It is understood that the content of the above method embodiments is applicable to the embodiments of this electronic device. The specific functions implemented by the embodiments of this electronic device are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0189] This application also provides a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement the random walk robot swarm control method as described in any of the above specific embodiments.

[0190] This application also discloses a computer program product, including a computer program or computer instructions, which are stored in a computer-readable storage medium. The processor of the computer device reads the computer program or computer instructions from the computer-readable storage medium and executes the computer program or computer instructions, causing the computer device to perform the random walk robot swarm control method as described in any of the preceding embodiments.

[0191] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0192] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, apparatus, product, or system that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatuses. It should be understood that in this application, “at least one” means one or more, and “more than one” means two or more.

[0193] In the several embodiments provided in this application, it should be understood that the disclosed apparatus, system, and method can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0194] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0195] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0196] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0197] Although the description of this application has been quite detailed and particularly focused on several of the described embodiments, it is not intended to limit itself to any of these details or embodiments or any particular embodiment. Rather, it should be considered as effectively covering the intended scope of this application by referring to the appended claims and taking into account the prior art, which provides for a broad possible interpretation of these claims. Furthermore, the foregoing description of this application with respect to embodiments foreseeable by the inventors is intended to provide a useful description, and non-substantial modifications to this application that have not yet been foreseen may still represent equivalent modifications.

Claims

1. A method for controlling a swarm of robots during random walks, characterized in that, The method includes: The robot swarm is divided into multiple robot groups, each of which includes multiple robots; Determine a virtual leader for each of the robot groups; For each of the robot groups, each robot moves with the virtual leader as its target; Each of the robot groups conducted its search using a random walk approach. Obtain the local location information of the robot group that discovered the target object, and assign each robot in the robot group that discovered the target object as a capture robot to perform a capture task against the target object; Based on the local location information, the area where the target object is located is divided into multiple grids; The grid corresponding to the location of the target object is marked as the first grid, and the grid corresponding to the location of the obstacle is marked as the second grid; Based on the location information and first distance of the first grid, a first concentration value is determined for each grid and a first concentration field is generated, wherein the first distance is the distance between the grids; Based on the location information and the second distance of the second grid, a second concentration value is determined for each grid and a second concentration field is generated, wherein the second distance is the distance between the grids; The first concentration field and the second concentration field are fused to obtain a fused third concentration field, and the third concentration value of each grid in the third concentration field is determined. Based on a preset safety distance and the location information of the target object located within the concentration field, an equipotential line of concentration value is generated around the target object; Each of the trapping robots is controlled to move closer to the target object, and the concentration value corresponding to the location of each trapping robot is obtained; When the concentration value corresponding to the location of the trapping robot is equal to the concentration value of the equipotential line, the trapping robot is controlled to stop moving.

2. The method for controlling a random walk robot swarm according to claim 1, characterized in that, The provision that, for each of the robot groups, each robot moves in response to the virtual leader includes: Each robot moves in the direction of a preset composite potential field, wherein the composite potential field includes a first potential field and a second potential field. When the robot detects the virtual leader, it generates an attraction based on the first potential field, causing the robot to move closer to the virtual leader under the influence of the attraction. When the robot detects a target object, it generates a repulsive force based on the second potential field, causing the robot to move away from the target object under the action of the repulsive force, wherein the target object is the virtual leader, the robot, or an obstacle.

3. The method for controlling a robot swarm during random walks according to claim 1, characterized in that, The method includes: When the robot group is conducting a search operation in the current area, the density of the robot group in the current area is determined based on the time interval between the occurrence of interference events among the multiple robot groups. When two robot groups performing search operations interfere with each other, it is recorded as an interference event. The time interval is the duration between the occurrence times of the two interference events.

4. The method for controlling a random walk robot swarm according to claim 3, characterized in that, The method further includes: The total average time interval is determined based on the multiple time intervals mentioned above; If the current time interval is greater than the total average time interval, then the step size of each robot in the robot group is reduced to narrow the search range of the robot group. If the current time interval is less than the total average time interval, the step size of each robot in the robot group is increased to expand the search range of the robot group.

5. The method for controlling a random walk robot swarm according to claim 1, characterized in that, The method includes: As the capture robot moves closer to the target object, the coordinate information of the capture robot is obtained; Based on the coordinate information of the encirclement robot, a first offset of the encirclement robot is determined, wherein the first offset is the coordinate offset between the encirclement robot and another encirclement robot; Based on the coordinate information of the target object and the coordinate information of the capture robot, a second offset of the capture robot is determined, wherein the second offset is the coordinate offset between the capture robot and the target object; Based on the first offset and the second offset, a third offset of the capture robot is determined; The movement of the capture robot is controlled based on the third offset.

6. A random walk robot swarm control system, characterized in that, The system includes a control device and a swarm of robots consisting of multiple robots; A grouping module is used to divide a group of robots into multiple robot groups, each of which includes multiple robots; The selection module is used to determine the virtual leader in each of the robot groups; A first control module is configured to target each of the robot groups, with each robot moving toward the virtual leader. The second control module is used to enable each robot group to perform search operations using a random walk approach; The third control module is used to acquire the local position information of the robot group that discovered the target object, and to make each robot in the robot group that discovered the target object act as a capture robot to perform the capture task against the target object; Based on the local location information, the area where the target object is located is divided into multiple grids; The grid corresponding to the location of the target object is marked as the first grid, and the grid corresponding to the location of the obstacle is marked as the second grid; Based on the location information and first distance of the first grid, a first concentration value is determined for each grid and a first concentration field is generated, wherein the first distance is the distance between the grids; Based on the location information and the second distance of the second grid, a second concentration value is determined for each grid and a second concentration field is generated, wherein the second distance is the distance between the grids; The first concentration field and the second concentration field are fused to obtain a fused third concentration field, and the third concentration value of each grid in the third concentration field is determined. Based on a preset safety distance and the location information of the target object located within the concentration field, an equipotential line of concentration value is generated around the target object; Each of the trapping robots is controlled to move closer to the target object, and the concentration value corresponding to the location of each trapping robot is obtained; When the concentration value corresponding to the location of the capture robot is equal to the concentration value of the equipotential line, control the capture robot to stop moving; Each of the robots is used to acquire distance information, coordinate information, and to identify target objects or obstacles.

7. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the random walk robot swarm control method according to any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the robot swarm control method for random walks as described in any one of claims 1 to 5.

Images