CN115113626B - Robot distributed control system and method thereof - Google Patents
Robot distributed control system and method thereof Download PDFInfo
- Publication number
- CN115113626B CN115113626B CN202210850293.7A CN202210850293A CN115113626B CN 115113626 B CN115113626 B CN 115113626B CN 202210850293 A CN202210850293 A CN 202210850293A CN 115113626 B CN115113626 B CN 115113626B
- Authority
- CN
- China
- Prior art keywords
- robot
- role
- task
- organizer
- robots
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0219—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory ensuring the processing of the whole working surface
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention relates to a robot distributed control system and a method thereof, belonging to the technical field of artificial intelligence, comprising the following steps: at least two robots, which have multiple roles: the explorer roles, the organizer roles and the executor roles are all started from the explorer roles to search for the target object, when the explorer role robots discover the target object, the explorer role robots are converted into the organizer roles, the rest explorer role robots are converted into the executor roles, and the organizer role robots organize the executor role robots to execute tasks on the target object; the main coordinator is communicated with the robot and monitors and judges the task state of the robot distributed control system, and if the robot distributed control system is in a deadlock state, the main coordinator coordinates the work task of the organizer role robot to enable the robot distributed control system to be in a normal operation state. The invention discloses a distributed robot control system which is used for organizing and coordinating a plurality of robots to cooperatively work, improving the working efficiency and completing the work tasks which can not be completed by a single robot.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a distributed control system and a distributed control method for a robot.
Background
In recent years, with the rapid development of artificial intelligence, particularly robot science, robots are gradually applied to various fields of society, people find that the operation capacity of a single robot is limited, so that the robot cannot complete and execute complex and various tasks.
Disclosure of Invention
The invention provides a distributed control system of robots and a method thereof, aiming at solving the problem that the existing single robot system cannot efficiently finish the work of complex diversity.
In order to solve the technical problems, the invention adopts the following technical scheme to realize:
in one aspect, an embodiment of the present invention provides a distributed robot control system, which includes:
at least two robots, which have multiple roles: explorer roles, organizer roles, actor roles;
the robots all start to search for the target object from explorer roles, when at least one explorer role robot finds the target object, the explorer role robot is converted into an organizer role, the other explorer role robots are converted into executor roles, and the organizer role robot organizes the executor role robot to execute tasks on the target object;
and the master coordinator is communicated with the robot and monitors and judges the task state of the robot distributed control system, and coordinates the work task of the organizer role robot if the master coordinator is in a deadlock state, so that the robot distributed control system is in a normal operation state.
In some embodiments, the robot is also provided with a proponentrole;
when at least one explorer role robot finds a target object, the explorer role robot is converted into the organizer role and carries out work task bidding, the rest explorer role robots are converted into the supporter role and participate in bidding of work tasks, the supporter role robot with the largest bidding utility value is selected to become an executor role, and the organizer role robot organizes the executor role robot to execute tasks on the target object.
In some embodiments, the bidding utility value is calculated by the bidding utility function U, and is specifically calculated as follows:
d is the shortest path distance to the position of the target object planned by the supporter role robot, a is the work task capacity of the supporter role robot, b is the work task amount to be completed after the work task capacity of the organizer role robot is removed, and k1 and k2 are bidding utility function coefficients.
In some embodiments, when the supporter role robot receives job task bid information issued by a plurality of organizer role robots, the supporter role robot calculates a task utility value of each bid job task by using a task utility function Ut, and takes the job task with the largest task function value to participate in bidding.
In some embodiments, the task utility function Ut is calculated as follows:
wherein d is the shortest path distance between the support role robot and the position of the target object, v is a task reward value, and m and n are task utility function coefficients.
In some embodiments, before the executor role robot starts to carry at the target object position, the executor role robot may also receive job task bid inviting information issued by other organizer role robots, calculate a task conversion utility value of each bid job task by using the task conversion utility function Qt, and participate in bidding for the job task with the task conversion utility value being the minimum positive value.
In some embodiments, the task conversion utility function Qt is specifically calculated as follows:
wherein d1 is the shortest path distance between the executor role robot and the position of the target object for planning and reaching the current task, v1 is the reward value of the current task, c1 is the times of the common conversion for executing the current task, d2 is the shortest path distance between the executor role robot and the position of the target object for planning and reaching the new task, v2 is the reward value of the new task, c2 is the times of the common conversion for executing the current task, lambda is the conversion utility function value coefficient, and m and n are the task utility function coefficients.
In some embodiments, the organizer role robot performs job task bidding within a predetermined range immediately after the executive role robot in the team switches tasks, and if no robot can satisfy the job task within the predetermined range, the predetermined range is expanded according to the ratio k to continue the job task bidding until a robot can satisfy the job task.
In some embodiments, the master coordinator comprises a blackboard structure unit, a monitoring unit, a coordinating unit;
the organizer role robot issues the distribution information of the organizer role robot to the work tasks on the blackboard structural unit, the monitoring unit regularly monitors and judges the work task bidding states of all the organizer role robots on the blackboard structural unit, if the monitoring unit finds that the work tasks of all the organizer role robots are in a bidding unsuccessful state twice, the monitoring unit judges that the robot distributed control system is in a deadlock state, then the coordination unit is called to coordinate the work tasks of the organizer role robots, the work tasks on the blackboard structural unit are changed, the robot distributed control system is changed into a normal operation state, and the distribution information of the work tasks on the blackboard structural unit is updated.
On the other hand, the embodiment of the invention provides a distributed control method for a robot, which comprises the following steps:
all robots initially search for a target object for an explorer role;
when at least one seeker role robot finds a target object, the robot is converted into an organizer role and carries out job task bidding, and the other seeker role robots are converted into supporters roles and participate in bidding of the job task;
selecting the supporter role robot with the maximum bidding utility value to become an executor role;
the organizer role robot organizes the task executed by the executor role robot on the target object and sends the distribution information of the organizer role robot on the work task to the master coordinator;
and the main coordinator receives the distribution information of the work tasks and judges the states of the work tasks, and if the work tasks are in a deadlock state, the main coordinator coordinates the work tasks to change the work tasks into a normal operation state.
Compared with the prior art, the distributed control system and the distributed control method for the robot have the technical effects that:
1. the invention adopts a multi-stage distributed cooperative system structure, which comprises a main coordinator only providing global information and at least two robots, wherein each robot can independently make a decision and can be used as a locally coordinated organizer role robot to organize other explorer role robots to work and interact with the main coordinator, so that a robot control system adopting the system structure is hierarchical in organization structure but completely distributed in control, and the robot control system can be used for organizing and coordinating a plurality of robots to cooperatively work, thereby improving the working efficiency and completing the work tasks which can not be completed by a single robot.
2. The organizer role robot of the robot distributed control system is used as an organizer of a work task to select a supporter role robot with the maximum competitive utility value to become an executor role to complete the work task, namely a small number of robots are used for completing complex tasks, so that a large amount of conflicts and waste of production resources are avoided, the task completion time is shortened, and the work efficiency is improved.
3. Before the executor role robot starts to execute tasks at the position of a target object, the executor role robot can also receive work task bidding information issued by other organizer role robots, calculate the task conversion utility value of each bidding work task by using a task conversion utility function Qt, and take the work task with the task conversion utility value of the minimum positive value to participate in bidding, so that the task completion time is shortened, and the system work efficiency is improved.
Drawings
FIG. 1 is a schematic diagram of a distributed control system for a robot according to the present invention;
FIG. 2 is a schematic diagram illustrating the role switching process of the robot according to the present invention;
FIG. 3 is a schematic view of the role transformation process of the robot with the role of the supporter according to the present invention;
FIG. 4 is a schematic structural diagram of a distributed control system of a robot including a blackboard structural unit according to the present invention;
FIG. 5 is a flow chart of a distributed control method of the robot of the present invention;
FIG. 6 is a schematic diagram of a distributed control system of a dual coordinator robot according to the present invention.
Detailed Description
In order to make those skilled in the art better understand the solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below with reference to the drawings in the embodiment of the present invention, and it is obvious that the described embodiment is only a part of the embodiment of the present invention, and not all embodiments. Other solutions, which can be obtained by a person skilled in the art without making creative efforts based on the embodiments of the present invention, shall fall within the protection scope of the present invention.
To facilitate understanding of embodiments of the present invention by those skilled in the art, technical terms related to the present invention are explained below.
The robot is an intelligent machine capable of working semi-autonomously or fully autonomously, has basic characteristics of perception, analysis, decision, execution and the like, can assist and even replace human beings to finish repeated, dangerous, heavy and complex work, improves the working efficiency and quality, serves human life, and expands or extends the activity and capability range of the human beings.
A multi-robot system refers to a collection of multiple robots having dynamic and kinematic characteristics and capable of interacting with other individuals.
Example 1
Fig. 1 shows a schematic structural diagram of a distributed control system for a robot according to the present invention, and it can be seen that the distributed control system for a robot comprises:
at least two robots, which have multiple roles: explorer roles, organizer roles, actor roles;
all robots start to search for the target object from the explorer roles, when at least one explorer role robot discovers the target object, the explorer role robot is converted into an organizer role, the other explorer role robots are converted into executor roles, and the organizer role robot organizes the executor role robot to execute tasks on the target object;
and the master coordinator is communicated with the robot and monitors and judges the task state of the robot distributed control system, mainly monitors and judges the task states of the explorer role robot, the organizer role robot and the executor role robot, and coordinates the work task of the organizer role robot if the explorer role robot, the organizer role robot and the executor role robot are in a deadlock state so that the robot distributed control system is in a normal operation state.
As shown in fig. 1, the robot team of the distributed robot control system of the embodiment may be composed of n (n is greater than or equal to 2) robots, and the number of the robots may be reasonably selected by the staff according to the actual work task requirement. The master coordinator and each robot can communicate in a wireless network, bluetooth, optical fiber and the like. The robot distributed control system is used for controlling the robot to perform tasks including carrying, catching, foraging, assembling, disassembling, exploring and the like.
As shown in fig. 1, the distributed control system for robots of this embodiment adopts a multi-level distributed cooperative control structure, in which a master coordinator only provides global information, and each robot as an independent individual can make an autonomous decision and can coordinate with each other.
As shown in fig. 2, each robot starts from the seeker role robot to search for the target object, when one robot finds the target object, it is converted into the organizer role robot, the other seeker role robots are converted into the executor role robots, the organizer role robot organizes the executor role robots to execute tasks on the target object, and in the above process, each robot is equal, and the robot that first found the target object is converted into the organizer role robot, so the control among the robots is distributed. Certainly, a plurality of robots find a plurality of target objects at the same time, the plurality of robots are converted into organizer role robots at the same time, a system work task deadlock state may occur when the plurality of organizer role robots organize the same executive role robot to execute tasks, and at the moment, the main coordinator coordinates the work tasks of the organizer role robots to enable the system work tasks to recover to a normal running state.
In the distributed control system of the robot, the main coordinator coordinates at least two robots to work cooperatively, so that the working efficiency is improved, and the work tasks which cannot be completed by a single robot can be completed.
In one embodiment of the invention, when a plurality of seeker role robots simultaneously discover the same target object, the seeker role robot with the largest task utility value is converted into an organizer role robot, and the organization executor role robot executes tasks on the target object. The task utility value is calculated by a task utility function T, and the specific calculation is as follows:
wherein di is the shortest path distance to the position of the target object planned by the explorer role robot or the explorer role robot within the preset range of the explorer role robot as the center, ai is the work task capacity of the explorer role robot or the explorer role robot within the preset range of the explorer role robot as the center, b is the work task amount to be completed by the target object, p1 and p2 are task utility function coefficients, and n is the number of the explorer role robots capable of participating in executing tasks on the target object within the preset range.
In this embodiment, when a plurality of seeker role robots simultaneously discover the same target object, each seeker role robot discovering the same target object is independent and immediately within a predetermined range, such as 1m, 2m, 5m, 10m, 50m, etc., a shortest path distance and a work task capability planned to reach the target object position by the seeker role robot capable of participating in the target object task are counted, the seeker role robot with the largest task utility value is converted into an organizer role robot through the task utility function T calculation, and the organizer role robot executes the task on the target object. According to empirical analysis, the working efficiency of the system is highest when the value range of p1 is (5, 10) and the value range of p2 is (20, 30). When the value range of n is (2, 4), the working efficiency of the system is highest.
Example 2
Fig. 3 is a schematic diagram illustrating a role conversion flow of the robot provided with a supporter role, and as can be seen from fig. 3, the robot of the embodiment is further provided with a supporter role;
when at least one explorer role robot finds the target object, the explorer role robot is converted into an organizer role and issues work task invitation information to perform work task invitation, the other explorer role robots are converted into supporters roles and participate in bidding of work tasks, the supporters role robot with the largest bidding utility value is selected to become an executor role, and the organizer role robot organizes the executor role robot to perform tasks on the target object. The bidding utility value is calculated by a bidding utility function U, and the calculation is as follows:
d is the shortest path distance to the position of the target object planned by the supporter role robot, a is the work task capacity of the supporter role robot, b is the work task amount to be completed after the work task capacity of the organizer role robot is removed, and k1 and k2 are bidding utility function coefficients.
In the bidding process of the organizer role robot, the supporter role robot with the largest utility value is always selected to complete the work task through calculating the bidding utility function U, namely, the organizer role robot completes the work task as independently as possible, if the work task amount is large, a small number of robots are selected to complete the work task as possible when a plurality of robots complete the work task cooperatively, so that a large number of conflicts are avoided, the coordination time among the robots is reduced, and the overall work efficiency is improved. According to empirical analysis, the system has the highest working efficiency when the value range of k1 is (2, 5) and the value range of k2 is (12, 17).
Example 3
In this embodiment, when the supporter role robot receives job task bid inviting information issued by a plurality of organizer role robots, the supporter role robot calculates a task utility value of each bid inviting job task by using a task utility function Ut, and takes the job task with the largest task function value to participate in bidding. The specific calculation mode of the task utility function Ut is as follows:
wherein d is the shortest path distance between the support role robot and the position of the target object, v is a task reward value, and m and n are task utility function coefficients.
In this embodiment, the supporter role robot calculates and evaluates each bidding task to participate in bidding by using the task utility function Ut, and selects the work task that can contribute the maximum performance to the system, i.e., the work task with the maximum task function value to participate in bidding, so as to reduce the invalid bidding behavior of the supporter role robot, shorten the completion time of the work task, and further improve the overall work efficiency. According to the experience, when the value range of m is (2, 5) and the value range of n is (30, 40), the working efficiency of the system is highest.
Example 4
In this embodiment, before the executor role robot starts to execute the task at the target position, the executor role robot may further receive job task bid inviting information issued by other organizer role robots, calculate a task conversion utility value of each bid inviting job task using the task conversion utility function Qt, and participate in bidding for the job task whose task conversion utility value is the minimum positive value. The specific calculation mode of the task conversion utility function Qt is as follows:
wherein d1 is the shortest path distance between the executor role robot and the position of the target object for planning and reaching the current task, v1 is the reward value of the current task, c1 is the times of the common conversion for executing the current task, d2 is the shortest path distance between the executor role robot and the position of the target object for planning and reaching the new task, v2 is the reward value of the new task, c2 is the times of the common conversion for executing the current task, lambda is the conversion utility function value coefficient, and m and n are the task utility function coefficients.
In this embodiment, the robot distributed control system has the highest working efficiency and the least time to complete a work task when the lambda value range is (0.6, 0.8) according to empirical analysis. In the embodiment, the running environment of the robot distributed control system is considered to be dynamically changed, the design is that as long as the executor role robot does not start to execute the task, the task conversion utility value of each bidding work task can be calculated through the task conversion utility function Qt, if the task conversion utility value is less than or equal to zero, the executor role robot still executes the original work task without performing the conversion of the work task, and if the task conversion utility value is greater than zero, the minimum positive value work task is selected to participate in bidding, so that the work task completion time is further shortened, and the system work efficiency is improved. The executor role robot receives the job task bid information issued by other organizer role robots when executing a certain job task, and can participate in the bidding of the job task again when the task conversion utility value is a positive value if the executor role robot performs a new job task in a conversion manner, so that the job task completion time is shortened on the whole, and the work efficiency of the system is improved.
In this embodiment, in order to reduce communication traffic of the robot distributed control system, after an organizer role robot switches tasks, it is designed that the organizer role robot does not use a global broadcast method to issue work task bidding information, but immediately performs work task bidding within a predetermined range, and if no robot within the predetermined range can satisfy the work task, the predetermined range is expanded according to a ratio k to continue work task bidding until a robot can satisfy the work task. The predetermined range is selected according to the requirement of actual work, for example, the predetermined range is covered by circles with the radius of 1m, 3m, 10m, 15m, 50m and the like, which are centered on the organizer role robot, and the predetermined range can also be a square area with flexibly set length and width. In the embodiment, the original task organizer role robot is designed to only carry out local bidding in a preset range, so that the traffic of the system can be reduced.
Example 5
Fig. 4 shows a schematic structural diagram of a distributed control system of a robot including a blackboard structural unit, and it can be known from the diagram that a master coordinator includes a blackboard structural unit, a monitoring unit, a coordinating unit, a storage unit, and an intelligent operation unit;
the organizer role robot issues the distribution information of the organizer role robot to the work tasks on the blackboard structural unit, the monitoring unit regularly monitors and judges the work task bidding states of all the organizer role robots on the blackboard structural unit, if the monitoring unit finds that the work tasks of all the organizer role robots are in a bidding unsuccessful state twice, the monitoring unit judges that the robot distributed control system is in a deadlock state, then the coordination unit is called to coordinate the work tasks of the organizer role robots, the work tasks on the blackboard structural unit are changed, the robot distributed control system is changed into a normal operation state, and the distribution information of the work tasks on the blackboard structural unit is updated.
In this embodiment, when all the organizer role robots are in the deadlock state in which bidding is unsuccessful, preferably, according to the unallocated task amounts of all the tasks, the task corresponding to the minimum unallocated task amount is selected to be preferentially coordinated, and by using the method similar to that in embodiment 4, a robot capable of performing the work task is selected within a predetermined range until the distributed control system of the robot is in the normal operation state, and finally, the allocation information of the work task is updated to the blackboard structure unit.
In the embodiment, a blackboard structure unit, a monitoring unit and a coordination unit are adopted only for providing global information, the work task allocation of each robot is coordinated only when the work task allocation has a deadlock state, namely, a global control mechanism does not exist, the robot distributed control system is hierarchical in the whole structure, but the control is completely distributed in a robot team, so that the robot distributed control system has high real-time performance, dynamic performance, fault tolerance and expandability.
As shown in fig. 6, in one embodiment of the present invention, the robot distributed control system has a master coordinator and a slave coordinator, and becomes a dual-coordinator robot distributed control system. The master coordinator and the slave coordinator are connected, preferably, the master coordinator and the slave coordinator are connected by a heartbeat line, and the slave coordinator and the robot are in communication through a network. The slave coordinator is a standby system of the master coordinator, the master coordinator periodically backs up information to the slave coordinator, or manually backs up information in the master coordinator to the slave coordinator, the slave coordinator detects the keep-alive state of the master coordinator and synchronously updates the information at regular time, in other aspects, the slave coordinator has the same function as the master coordinator, when the master coordinator fails or is attacked and fails, the slave coordinator temporarily takes over the work of the master coordinator, so that the robot distributed control system continues to operate normally, once the slave coordinator detects that the master coordinator is recovered to be normal, the state of the slave coordinator is converted back to be subordinate, the master coordinator is recovered to be normal, and starts to execute tasks, so that the anti-interference and anti-interference capability of the system is improved, and the system has better stability.
Example 6
Fig. 5 shows a flow chart of the distributed control method of the robot of the invention, and it can be known that the distributed control method of the robot comprises the following steps:
s1, all robots initially search for the target object for the role of an explorer.
When the distributed control system of the robot is started, all the robots are initially in the role of an explorer to search for a target object.
S2, when at least one explorer role robot finds the target object, the explorer role robot is converted into an organizer role and carries out work task bidding, and the other explorer role robots are converted into supporters roles and participate in bidding of work tasks.
When the supporter role robot receives the job task bid inviting information issued by the plurality of organizer role robots, the supporter role robot preferably calculates the task utility value of each bid inviting job task by using a task utility function Ut, and takes the job task with the maximum task function value to participate in bidding. The specific calculation mode of the task utility function Ut is as follows:
wherein d is the shortest path distance between the support role robot and the position where the support role robot plans to reach the target object, v is a task reward value, and m and n are task utility function coefficients.
Preferably, the supporter role robot uses the task utility function Ut to calculate and evaluate each bidding task, and takes the work task which can contribute the maximum performance to the system, i.e. the work task with the maximum task function value, to participate in bidding, so as to reduce the invalid bidding behavior of the supporter role robot, shorten the completion time of the work task, and further improve the overall work efficiency.
And S3, selecting the supporter role robot with the maximum bidding utility value to become an executor role.
The bidding utility value is calculated by a bidding utility function U, and is specifically calculated as follows:
d is the shortest path distance to the position of the target object planned by the supporter role robot, a is the work task capacity of the supporter role robot, b is the work task amount to be completed after the work task capacity of the organizer role robot is removed, and k1 and k2 are bidding utility function coefficients.
In the bidding process, the organizer role robot calculates through a bidding utility function U, and selects the supporter role robot with the largest utility value all the time to complete the work task, namely the organizer role robot completes the work task as independently as possible, and if the work task amount is large, a small number of robots are selected as possible to complete the work task when a plurality of robots cooperate to complete the work task, so that a large number of conflicts are avoided, the coordination time among the robots is reduced, and the overall work efficiency is improved.
And S4, the organizer role robot organizes the executive role robot to execute the task on the target object and sends the distribution information of the executive role robot to the work task to the main coordinator.
And S5, the main coordinator receives the distribution information of the work tasks and judges the states of the work tasks, and if the work tasks are in a deadlock state, the main coordinator coordinates the work tasks to enable the work tasks to be in a normal operation state.
The master coordinator receives the distribution information of the work tasks of the organizer role robots and regularly monitors and judges the work task bidding states of all the organizer role robots, if the master coordinator finds that the work tasks of all the organizer role robots are unsuccessfully bidding states twice, the master coordinator judges that the robot distributed control system is in a deadlock state, and further coordinates the work tasks of the organizer role robots, so that the robot distributed control system is in a normal operation state, and the distribution information of the work tasks of all the organizer role robots is updated.
In this embodiment, when all the organizer role robots are in the deadlock state in which bidding is unsuccessful, preferably, according to the unallocated task amounts of all the tasks, a task corresponding to the minimum unallocated task amount is selected to be preferentially coordinated, and by using a method similar to that in embodiment 4, a robot capable of meeting the work task is selected within a predetermined range until the whole robot system is in a normal operation state, and finally, the allocation information of the work task is updated.
In the embodiment, the work task allocation of each robot is coordinated only when the work task allocation has a deadlock state, that is, a global control method does not exist, and the robot distributed control method completely distributes the control right in a robot team, so that a system applying the robot distributed control method has high real-time performance, dynamic performance, fault tolerance and expandability.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (10)
1. A distributed control system for a robot, comprising:
at least two robots, which have multiple roles: seeker roles, organizer roles, executor roles;
the robots all start to search for the target object from explorer roles, when at least one explorer role robot finds the target object, the explorer role robot is converted into an organizer role, the other explorer role robots are converted into executor roles, and the organizer role robot organizes the executor role robot to execute tasks on the target object;
when a plurality of seeker role robots find the same target object at the same time, the seeker role robot with the maximum task utility value is converted into an organizer role robot, and an organizer role robot is organized to execute tasks on the target object;
and the master coordinator is communicated with the robot and monitors and judges the task state of the robot distributed control system, and coordinates the work task of the organizer role robot if the master coordinator is in a deadlock state, so that the robot distributed control system is in a normal operation state.
2. The distributed control system of robots as recited in claim 1 wherein said robot is further provided with a supporter role;
when at least one explorer role robot finds a target object, the explorer role robot is converted into the organizer role and carries out work task bidding, the rest explorer role robots are converted into the supporter roles and participate in bidding of work tasks, the supporter role robot with the largest bidding utility value is selected to become an executor role, and the organizer role robot organizes the executor role robot to execute tasks on the target object.
3. The distributed robot control system of claim 2, wherein the competitive utility value is calculated from a competitive utility function U, specifically as follows:
wherein d is the shortest path distance to the position of the target object planned by the supporter role robot, a is the work task capacity of the supporter role robot, b is the work task amount to be completed after the work task capacity of the organizer role robot is removed, and k is the distance between the supporter role robot and the target object 1 、k 2 Are the bidding utility function coefficients.
4. The distributed robot control system of claim 2, wherein when the supporter role robot receives job task bid information issued by a plurality of the organizer role robots, it calculates a task utility value of each bid job task using a task utility function Ut, and takes the job task having the largest task function value to participate in bidding.
5. The distributed robot control system of claim 4, wherein the task utility function Ut is calculated as follows:
wherein d is the shortest path distance between the support role robot and the position of the target object, v is a task reward value, and m and n are task utility function coefficients.
6. The distributed robot control system of claim 4, wherein the actor role robot receives job task bid information issued by other organizer role robots before starting to execute a task at the target location, calculates a task conversion utility value for each bid job using the task conversion utility function Qt, and participates in bidding for the job task with the least positive task conversion utility value.
7. The distributed robot control system of claim 6, wherein the task conversion utility function Qt is calculated in the following manner:
wherein d is 1 Planning the shortest path distance to the target position of the current task for the actor character robot 1 Awarding a value for the current task, c 1 Number of times conversion co-occurs for executing the current task, d 2 Planning the shortest path distance, v, for the actor character robot to the target location of the new task 2 Awarding values for new tasks, c 2 In order to execute the times of the common conversion of the current task, lambda is a conversion utility function value coefficient, and m and n are task utility function coefficients.
8. A distributed robot control system as in claim 6 wherein said organizer role robot performs job task bidding within a predetermined range immediately after a task is switched by its executive role robots in its team, and if no robot can satisfy the job task within said predetermined range, continues job task bidding by enlarging the predetermined range by a ratio k until there is a robot that can satisfy the job task.
9. A distributed control system of robots according to any of the claims 1 to 8, characterized in that the master coordinator comprises a blackboard construction unit, a monitoring unit, a coordination unit;
the organizer role robot issues distribution information of the organizer role robot to work tasks on the blackboard structural unit, the monitoring unit regularly monitors and judges all work task bidding states of the organizer role robot on the blackboard structural unit, if the monitoring unit finds that all work tasks of the organizer role robot are in a bidding unsuccessful state for two times, the robot distributed control system is judged to be in a deadlock state, then the coordination unit is called to coordinate the work tasks of the organizer role robot, the work tasks on the blackboard structural unit are changed, the robot distributed control system is changed into a normal operation state, and the distribution information of the work tasks on the blackboard structural unit is updated.
10. A distributed control method for a robot is characterized by comprising the following steps:
all robots initially search for the target object for the role of an explorer;
when at least one seeker role robot finds a target object, the robot is converted into an organizer role and carries out job task bidding, and the other seeker role robots are converted into supporters roles and participate in bidding of the job task;
when a plurality of seeker role robots find the same target object at the same time, the seeker role robot with the maximum task utility value is converted into an organizer role robot, and an organizer role robot is organized to execute tasks on the target object;
selecting the supporter role robot with the maximum bidding utility value to become an executor role;
the organizer role robot organizes the executive role robot to execute tasks on the target object and sends the distribution information of the executive role robot to the work task to the main coordinator;
and the main coordinator receives the distribution information of the work tasks and judges the states of the work tasks, and if the work tasks are in a deadlock state, the main coordinator coordinates the work tasks to enable the work tasks to be in a normal operation state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210850293.7A CN115113626B (en) | 2022-07-20 | 2022-07-20 | Robot distributed control system and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210850293.7A CN115113626B (en) | 2022-07-20 | 2022-07-20 | Robot distributed control system and method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115113626A CN115113626A (en) | 2022-09-27 |
| CN115113626B true CN115113626B (en) | 2023-04-11 |
Family
ID=83331594
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210850293.7A Active CN115113626B (en) | 2022-07-20 | 2022-07-20 | Robot distributed control system and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115113626B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116185044B (en) * | 2023-04-26 | 2023-06-27 | 威康(深圳)智能有限公司 | Control method, device, equipment and system of robot cluster system |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1487871A (en) * | 2001-10-22 | 2004-04-07 | 索尼公司 | Robot apparatus and control method thereof |
| CN1877475A (en) * | 2005-06-09 | 2006-12-13 | 索尼株式会社 | Network system, mobile device, method of controlling same, and computer program |
| CN102033536A (en) * | 2010-12-22 | 2011-04-27 | 大连理工大学 | Scheduling, organization and cooperation system and method for multi-robot system |
| CN105843227A (en) * | 2016-04-15 | 2016-08-10 | 上海大学 | Task intensity dynamic adjustment based multiple robots cooperating task hunting allocation algorithm |
| KR20190011856A (en) * | 2017-07-25 | 2019-02-08 | 전자부품연구원 | Work distribution device for efficient robot operation, system and method using the same |
| CN109814556A (en) * | 2019-01-22 | 2019-05-28 | 东南大学 | A kind of multi-robot Cooperation explores the apparatus and method of circumstances not known and map structuring |
| CN111240319A (en) * | 2019-12-31 | 2020-06-05 | 南京理工大学 | Outdoor multi-robot cooperative operation system and method thereof |
| CA3035556A1 (en) * | 2019-03-04 | 2020-09-04 | Von's Buffet Inc. | A motion interpretation system |
| CN111633655A (en) * | 2020-06-06 | 2020-09-08 | 杭州电子科技大学 | Traffic scheduling method for distributed autonomous mobile robot |
| CN111931384A (en) * | 2020-09-01 | 2020-11-13 | 中国人民解放军国防科技大学 | Group cooperative trapping method based on antenna model and storage medium |
| CN112671987A (en) * | 2021-01-22 | 2021-04-16 | 王海权 | Emergency rescue calling system |
| CN114336980A (en) * | 2022-03-07 | 2022-04-12 | 威康(深圳)智能有限公司 | Intelligent cloud power distribution method and system based on instruction transmission |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI333178B (en) * | 2007-07-13 | 2010-11-11 | Ind Tech Res Inst | Method for coordinating cooperative robots |
| WO2018163174A1 (en) * | 2017-03-07 | 2018-09-13 | B. G. Negev Technologies And Applications Ltd., At Ben-Gurion University | Market equilibrium mechanism for task allocation |
| CN113671987B (en) * | 2021-07-29 | 2024-02-02 | 北京理工大学 | Multi-machine distributed time sequence task allocation method based on non-deadlock contract net algorithm |
-
2022
- 2022-07-20 CN CN202210850293.7A patent/CN115113626B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1487871A (en) * | 2001-10-22 | 2004-04-07 | 索尼公司 | Robot apparatus and control method thereof |
| CN1877475A (en) * | 2005-06-09 | 2006-12-13 | 索尼株式会社 | Network system, mobile device, method of controlling same, and computer program |
| CN102033536A (en) * | 2010-12-22 | 2011-04-27 | 大连理工大学 | Scheduling, organization and cooperation system and method for multi-robot system |
| CN105843227A (en) * | 2016-04-15 | 2016-08-10 | 上海大学 | Task intensity dynamic adjustment based multiple robots cooperating task hunting allocation algorithm |
| KR20190011856A (en) * | 2017-07-25 | 2019-02-08 | 전자부품연구원 | Work distribution device for efficient robot operation, system and method using the same |
| CN109814556A (en) * | 2019-01-22 | 2019-05-28 | 东南大学 | A kind of multi-robot Cooperation explores the apparatus and method of circumstances not known and map structuring |
| CA3035556A1 (en) * | 2019-03-04 | 2020-09-04 | Von's Buffet Inc. | A motion interpretation system |
| CN111240319A (en) * | 2019-12-31 | 2020-06-05 | 南京理工大学 | Outdoor multi-robot cooperative operation system and method thereof |
| CN111633655A (en) * | 2020-06-06 | 2020-09-08 | 杭州电子科技大学 | Traffic scheduling method for distributed autonomous mobile robot |
| CN111931384A (en) * | 2020-09-01 | 2020-11-13 | 中国人民解放军国防科技大学 | Group cooperative trapping method based on antenna model and storage medium |
| CN112671987A (en) * | 2021-01-22 | 2021-04-16 | 王海权 | Emergency rescue calling system |
| CN114336980A (en) * | 2022-03-07 | 2022-04-12 | 威康(深圳)智能有限公司 | Intelligent cloud power distribution method and system based on instruction transmission |
Non-Patent Citations (1)
| Title |
|---|
| 胡俊 ; 朱庆保 ; .基于动态预测目标轨迹和围捕点的多机器人围捕算法.电子学报.2011,第30卷(第11期),第2481-2485页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115113626A (en) | 2022-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhao et al. | A heuristic distributed task allocation method for multivehicle multitask problems and its application to search and rescue scenario | |
| CN111465475B (en) | Method for operating a robot in a multi-agent system, robot and multi-agent system | |
| CN102033536B (en) | Scheduling, organization and cooperation system and method for multi-robot system | |
| CN115113626B (en) | Robot distributed control system and method thereof | |
| CN109814556A (en) | A kind of multi-robot Cooperation explores the apparatus and method of circumstances not known and map structuring | |
| CN110443412A (en) | The intensified learning method of Logistic Scheduling and path planning in dynamic optimization process | |
| CN109116854A (en) | A kind of robot cooperated control method of multiple groups based on intensified learning and control system | |
| CN107831685B (en) | Group robot control method and system | |
| CN112180734A (en) | Multi-agent consistency method based on distributed adaptive event triggering | |
| CN110070235A (en) | A kind of flexible dispatching method of multiple mobile robot | |
| CN108985580B (en) | Multi-robot disaster search and rescue task allocation method based on improved BP neural network | |
| CN111784211B (en) | Cluster-based group multitask allocation method and storage medium | |
| CN111798097B (en) | Autonomous mobile robot task allocation processing method based on market mechanism | |
| Ghavamzadeh et al. | Continuous-time hierarchical reinforcement learning | |
| CN112731942A (en) | Multi-AUV formation control method based on improved navigator virtual structure method | |
| CN108415760A (en) | A kind of online method for allocating tasks of gunz calculating based on mobile opportunistic network | |
| CN112101773A (en) | Task scheduling method and system for multi-agent system in process industry | |
| Ahn et al. | Coordination of two robotic manipulators for object retrieval in clutter | |
| Zhu et al. | Deep reinforcement learning for fairness in distributed robotic multi-type resource allocation | |
| CN109886574B (en) | Multi-robot task allocation method based on improved threshold method | |
| Camelo et al. | A scalable parallel Q-learning algorithm for resource constrained decentralized computing environments | |
| Sudacevschi et al. | Distributed Decision-Making Multi-Agent System in Multi-Dimensional Environment | |
| CN110266515A (en) | A kind of operation information system based on general fit calculation | |
| CN112184400A (en) | Heterogeneous multi-agent multi-stage distributed auction algorithm based on local information | |
| CN113635302B (en) | Integrated mobile cooperative robot control system based on field bus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |