KR101152251B1 - Multi robots control system and method of control the multi robots - Google Patents

Abstract

PURPOSE: A system and a method for controlling multiple robots are provided to set temporal restriction conditions for unit missions, to enable the unit missions based on that, and to enable a proper operator to intervene if necessary. CONSTITUTION: A system for controlling multiple robots comprises multiple robots, a network, and a server. The multiple robots respectively implement multiple action missions. The network provides action information of the multiple robots and paths for data communication. The server plans missions that the multiple robots implement to be coincided with restriction conditions according to each mission, and manages through data communications using the network in real time. The restriction conditions for the unit missions that the multiple robots implement include places and temporal restriction conditions including a start time, a finish tine, and a duration time.

Description

Multi robots control system and method of control the multi robots}

The present invention relates to a control system and a control method for operating multiple robots, and more particularly, to a control system and a control method capable of constituting a task of a multi-robot and optimizing management of the task performance.

In general, an autonomous robot control software architecture is a hybrid of a deliberate control software for intelligence and a reactive control software for autonomous responsiveness. It has a 3-tier control architecture. However, the field of robot control software structure that has been carried out so far focuses on the technology of the abstract structure and has not yet disclosed the specific implementation method.

Middleware technology for decentralized software integration has, to date, been substantially developed for enterprise-level computer software, such as parallel computer technology. Therefore, the contents of guaranteeing the real-time required by a system such as a large part of the robot are omitted, and only the prior art related to the independent communication middleware or the independent platform middleware exists, and the invention regarding the middleware comprehensively covering the entire layer is still It has not been disclosed.

Recently, as an example of applying a decentralized software integration technology to a robot, an object of a famous pet robot of patent 2000-7015057 (Applicant: Sony Kabushiki Kaisha, Idei Nobuyuki) developed in Japan and filed in Korea ( There is an invention related to a communication method between objects).

Robot operation is possible by proximity control, remote control and autonomy. Among them, autonomous robot operation requires the creation of a task and its management method so that the robot can perform its assigned task autonomously and continuously, and appropriate operator intervention should be possible if necessary.

The task assigned to the robot may consist of a single action such as movement or monitoring, but may be combined with a plurality of actions having different characteristics when necessary. For example, monitoring is performed after moving to a specific point. However, a task planning method has not yet been developed for robots to autonomously perform a plurality of actions having different performance characteristics, not just one action. The same is true for the management of the mission plan. In addition, even when the robot autonomously performs a task, appropriate operator intervention is required, but a solution for this has not been suggested in detail.

1 is an exemplary view showing the control of a multi-robot according to the prior art, showing the task plan of the multi-robot as a form utilizing a unit mission. Mission 101 identifies and defines its components 102, i.e., unit tasks in advance, and sets the constraints (start time, end time, and duration) for each unit mission over time. Includes the ability to plan the overall mission.

However, according to the task depiction method of FIG. 1, there is a disadvantage in that where the unit task is performed and the interrelationship between the unit tasks cannot be expressed.

It is an object of the present invention to provide a multi-robot control system and a multi-robot control method capable of setting and managing a plan so that a robot can perform a task autonomously and continuously.

Another object of the present invention is to provide a multi-robot control system and a multi-robot control method for setting time constraints on a unit task and enabling task management of the unit task based on the time constraint.

Another object of the present invention is to provide a multi-robot control system and a multi-robot control method capable of appropriate operator intervention as necessary.

 A multi-robot control system according to the present invention for achieving these objects is a plurality of robots each performing a plurality of operation tasks; A network providing a path for data communication and motion information of the plurality of robots; And a server configured to perform a task to be performed by the plurality of robots in accordance with constraints for each mission and to perform management in real time through data communication using the network.

A feature of another configuration of the multi-robot control system according to the present invention is that the constraints on the unit task performed by the plurality of robots include location and temporal constraints, and the temporal constraints include start time, end time, and duration. It includes a point.

A feature of another configuration of the multi-robot control system according to the invention is that the time constraints can be set in whole or in part when planning the robot's task in the server.

Another configuration of the multi-robot control system according to the present invention is that the server in real time in consideration of the position and movement of the robot in each unit time interval in the robot operation task for the occurrence of inconsistency and sudden situation for each unit mission It is management.

Another feature of the multi-robot control system according to the invention is that the server can add waypoints / missions / missions in real time to the robot's operational mission.

Another feature of the multi-robot control system according to the present invention is that when the server is inconsistent with the constraints on the performance of the robot's task, deferment, curtailment, cancel, stand by -by, by selecting any one of the control signal (pull) and extension (extension) to manage the robot in real time.

The multi-robot control method according to the present invention comprises the steps of planning a task by the network configuration using N unit tasks for a plurality of robots; Assigning tasks to the plurality of robots; It is characterized in that it comprises a process of managing the task of each robot manually or automatically to solve the inconsistency between the tasks performed by a plurality of robots in preparation for the planned task.

Another feature of the multi-robot control method according to the present invention comprises the steps of the task planning step, the task planning network, using a starting point / target point / waypoint / mission point / mission to plan the task; The method further includes a method of setting the time constraints (start time, end time, duration) in whole or in part for each unit mission.

Another feature of the multi-robot control method according to the present invention is that the process of assigning a task to a plurality of robots allocates an optimal task in consideration of the dependency relationship, the robot collaboration, and the task load.

Another feature of the multi-robot control method according to the present invention is that the process of assigning a task in consideration of the above-mentioned relationship is m _i is the i-th unit mission, m _{i '} is the successor mission of the i-th unit mission, t ^e _i is m the end of the time _i, t ^s _{i 'is} a m _i' to indicate the start time of the, in the presence of a relationship of _{'m i>} m _{i between'} m _i and m _i, t _i ^{e s} _i ≤t _'is satisfied Assign missions whenever possible.

Another feature of the multi-robot control method according to the present invention is that the process of assigning a task in consideration of the robot collaboration is to assign a unit task of a robot to be input to perform the task divided into forced collaboration and autonomous collaboration according to a given condition. to be.

Another feature of the multiple-robot control method according to the invention the forced collaboration t ^s _{i ^'s} ≤t ≤t _i ^e _i' or This point is assigned when the condition of t ^s _{i '} ≤ t ^e _i ≤ t ^e _{i' and} the condition of X _ij + X _{i'j '} ≥ 2 are satisfied.

The multi-robot control system and the multi-robot control method according to the present invention may have the following effects.

First, in the autonomous and continuous operation of robots, it is possible to plan complex tasks that are mixed with various performance characteristics.

Second, if the robot is inconsistent with a given task, it can be managed automatically or manually.

Third, it is possible to reschedule missions for changes in factors to be considered between robot operations.

1 is a unit configuration based on the unit task of a multi-robot according to the prior art.
2 is an exemplary diagram of a multi-robot control according to an embodiment of the present invention.
3 is a flow chart showing the progress of the multi-robot control method according to the present invention.
Figure 4 is a block diagram showing the configuration of a multi-robot control system according to the present invention.
5 is a conceptual diagram of the task management by the time constraints set in the unit mission.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described herein and may be embodied in other forms. Rather, the embodiments introduced herein are provided so that the disclosed contents can be thorough and complete, and to fully convey the spirit of the present invention to those skilled in the art. Like reference numerals denote like elements throughout the specification, and duplicate description thereof will be omitted.

The task performance by the control method according to the present invention for overcoming the disadvantages of FIG. 1 is performed as in FIG. 2. At this time, the starting point 201 and the final target point 207 from which the robot first starts, and a waypoint 202 (WP: Way Point) to pass through for performing a mission therebetween; The unit mission is performed between the mission point (203, MP: Mission Point, the point or place where the unit mission is performed while the robot is fixed) and the mission line (204, ML: Mission Line, the movement of the robot). To be able to plan missions by constructing a network.

According to the present invention, the task planning method using the network configuration as shown in the example of FIG. 2 can express information about a place where each unit task is performed, as well as the interrelationship between the unit tasks (such as after-sales relations).

As an example of the mission plan shown in FIG. 2, the robot moves from the starting point 201 to the gathering place 202 (unit mission # 1) and moves to the mission start line 203 according to the command (unit mission # 3). )do. The robot then performs a submission (unit mission # 4, eg, "road reconnaissance") between movements from the start line to subsequent mission points, and the main mission (unit mission # 5, eg " Surveillance "). After that, carry out the main task between maneuvers (unit mission # 7, eg "fire") up to the target point at the scheduled time. Sub-missions (Unit mission # 6, yes) to support the mission (unit mission # 2, eg "conveyance of goods") from the starting point to waypoint 206 and the main mission (unit mission # 7) to support mission performance. For example, "target acquisition" is additionally performed.

With reference to FIG. 2, the configuration of a mission may be established between the mission line, the mission point, and the waypoint (hereinafter, referred to as a "node") in consideration of the direction of execution of the mission. Otherwise, it is possible to assign the unit task to another robot before the start or end time of the unit task to be performed at the node (eg, 203). In this case, a new mission line may be generated by generating a location of the robot as a new node. However, as shown in FIG. 2, the method of depicting a task also has a disadvantage in that it cannot express information on time constraints (start time / end time / continuous time) for a unit task as in the related art (FIG. 1). In addition, the object includes, according to the present invention, for the manual or automatic task management of the planned task, to set the time constraints on the unit task and to enable the task management of the unit task based on it.

3 is a flowchart illustrating a process of a control method using multiple robots according to the present invention.

Using the N unit tasks for a number of robots to plan the task by the network configuration (S301), assigning the task to the plurality of robots according to the plan (S302), a plurality of robots in preparation for the planned task In order to solve the inconsistency between the tasks performed by the manual or automatic management of the task of each robot is made to include a step (S303).

Figure 4 is a block diagram showing the configuration of a multi-robot control system according to the present invention.

A network 20 which provides a path for communication of operation information and data of the plurality of robots 11, 12, 13..., 1N, and the plurality of robots 11, 12, 13 ... ; The server 30 to plan the task to be performed by the plurality of robots (11, 12, 13 ... 1N) to meet the constraints for each mission and to perform management in real time through data communication using the network 20 It is made, including.

Constraints on unit tasks performed by the plurality of robots 11, 12, 13..., 1N include location and temporal constraints, and the temporal constraints include start time, end time, and duration. The temporal constraint may be set in whole or in part when planning the task of the robot at the server.

The server 30 manages in real time in consideration of the position and movement of the robot for each unit time interval in the robot operation task for the occurrence of inconsistency and unexpected situation for each unit task. In addition, the server 30 may add the waypoint / mission / mission in real time to the operation task of the robot. When the server 30 performs inconsistency with respect to the constraints on the performance of the robot's task, the server 30 may perform deferment, curtailment, cancellation, stand-by, pull, and extension. ) To control the robot in real time.

5 is a conceptual diagram of the task management by the time constraints set in the unit mission. In order to explain the mission management method in detail, related symbols and symbols are defined as follows.

m _i : Unit mission i (i = 1, 2,… N)

t ^s _i : start time of m _i

t ^e _i : End time of m _i

: m_i의 지속시각, 즉 t^e _i - t^s _i

is the duration of m _i , t ^e _i -t ^s _i

: m_i의 최대 가능 지속시간

: Maximum possible duration of m _i

: m_i의 최소 요구 지속시간

: minimum required duration of m _i

>: The relationship between the two missions, i.e. m _i m _{i '} means m _i precedes m _i'

0: no setting

For each unit mission, the temporal constraints (start time, end time, duration) can be set in whole or in part during the task construction phase. In addition, as the unit missions 301 m _i and m _{i ′} (m _i > m _{i ′} ) are performed, inconsistencies with respect to the constraints set in m _i may occur, and the occurrence may be predicted in advance.

One of the following six mission management methods can be applied to m _i for automatic mission management against inconsistencies that occur.

가 불변적인 경우에 m_i 순연(303)? Deferment: In the situation 302 where m _i is delayed against the plan, i.e. t ^e _i > t ^s _{i '} , t ^s _i = t ^e _{i '} is established

M _i purely (303)

≠ 0일 때 (t^e _i-t^s _i) ≥

인 경우에 mi단축(304)? Curtailment: t ^e _i > t ^s _{i '} , t ^s _i = t ^e _{i '} is established and t ^e _i is immutable; Or in addition

≠ 0 (t ^e _i -t ^s _i ) ≥

If mi shortened (304)

≠ 0일 때 (t^e _i-t^s _i) <

인 경우에 취소(305)? Cancel: t ^e _i > t ^s _{i '} , t ^s _i = t ^e _{i '} is not established or t ^s _i = t ^e _i is established, t ^e _i is immutable

≠ 0 (t ^e _i -t ^s _i ) <

Cancel if is (305)

? Standby: Standby 305 in situations where m _i terminates prematurely, i.e. t ^e _i <t ^s _i , t ^s _i = t ^e _i

가 불변적인 경우에 m_i견인(308)? Pull: when t ^e _i <t ^s _i , t ^s _i = t ^e _i is established, t ^e _i is variable

M _i is the tow 308 immutable if

≠ 0일 때 (t^e _i-t^s _i) ≤

인 를 만족하는 새로운 t^s _i까지 m_i 신장(309)? Extension: when t ^e _i <t ^s _i and t ^s _i is variable and t ^e _i is invariant, tow to t ^s _i = t ^e _i ;

≠ 0 (t ^e _i -t ^s _i ) ≤

M _i up to new t ^s _i that satisfies (309)

In addition, the task management method includes the same for m _i and m _{i '} (m _i > m _i' ), which may be applied to m _i being performed instead of subsequent m _i .

In addition, the six task management methods may be implemented by simply prioritizing time and duration, in addition to implementation according to input of time and duration related constraints. As an example therefor, when t ^e _i > t ^s _i , if the time takes precedence over the duration “substantiation” is applied; Conversely, if the duration takes precedence over time, the "shortening" applies.

R _j : robot j ()

X _ij : Mission assignment determinant, 1 if the task is assigned to R _j , 0 otherwise.

L _j : Mission load of robot j

L ^max _j : The maximum mission load robot j can tolerate

The task assignment method may be solved by considering the post-relationship relationship, robot cooperation, and task load existing between unit tasks in assigning a plurality of robots to a task composed of N unit tasks.

In detail, the method of considering the propriety relation should be assigned a task such that t ^e _i ≤ t ^s _{i '} is satisfied when a relationship m _i > m _i' exists between m _i and m _{i '} .

In detail, how to consider the robot collaboration, the mandatory collaboration that requires the cooperation by two or more robots to perform a certain m _i and m _{i '} is based on the m _i (1) ) And Eq. (2) must be satisfied.

t ^s _{i ^'s} ≤t ≤t ^e _{i i'} or _{^{t s i '≤t e i ≤t}} e i' ... … … … … … … … … … … … … (One)

X _ij + X _{i'j '>} 2... … … … … … … … … … … … … … … … … … … … … (2)

Although not compulsory collaboration, autonomous collaboration in which a unit task is divided into K unit missions ( k = 1, 2, ... K ) for other reasons must be satisfied by equation (3).

………………………………………………………………(3)

… … … … … … … … … … … … … … … … … … … … … … … … (3)

In addition, for the autonomous collaboration method, m _k is m as an effect of collaboration that occurs in connection with the number of robots ( K ) input to collaboration when ( k = 1, 2, ... K ) is simultaneously performed. The duration compared to _i can be shortened as in Equation (4) below. This includes the application of various forms of mission load coordination and distribution in assignment.

……………………………………………………………(4)

… … … … … … … … … … … … … … … … … … … … … … … (4)

For the method of considering mission load, the task load assigned to robot j can be estimated as the duration of the task and it must be below the level that can be tolerated, and for this, Eq. (5) must be satisfied.

…………………………………………(5)

… … … … … … … … … … … … … … … … (5)

In addition, the task load equalization of Equation (6) and the task duration of Equation (7) are applied to the task assignment method of the robot in order to solve the inconsistency and increase the possibility of completing the task in connection with the task management method. It can optimize the assignment of robots in real time, taking into account minimization and predictive mission readjustment.

또는

……………………………(6)

or

… … … … … … … … … … … (6)

…………………………………………………………………(7)

… … … … … … … … … … … … … … … … … … … … … … … … … (7)

는 각각 전체 로봇에 대한 평균적 임무부하로서 식 (8)과 같이 정의될 수 있음을 포함한다.Of the formula (6)

Each contains the average workload for the entire robot, which can be defined as Eq. (8).

………………………………………………………………(8)

… … … … … … … … … … … … … … … … … … … … … … … … (8)

In addition, in the task assignment method of the robot, in order to solve the inconsistency between the task execution and increase the possibility of completion of the task in connection with the task management method, the entire task execution time may be divided and managed by unit time; .

Inconsistency (frequency and its probabilistic distribution characteristics) for each unit mission and various incidents (events or changes that cause disruptions to the planned task, such as failures, bullets / strucks, damages, disruptions / disruptions, malfunctions / losses) (Frequency and distribution characteristics) For the occurrence, it is possible to add waypoints / missions / missions if necessary. In addition, the task assignment may be optimized in real time in consideration of the position / movement of the robot for each unit time interval (time frame).

Applying the task plan according to the network configuration of FIG. 2 in the real time task assignment optimization; Probabilistic methods, such as simulations that take into account the frequency and distribution characteristics associated with inconsistencies and outbreaks of unit missions, are included.

Claims (12)

A plurality of robots each performing a plurality of operational tasks;
A network providing a path for data communication and motion information of the plurality of robots;
It comprises a server to plan the tasks to be performed by the plurality of robots to meet the constraints for each mission and to perform management in real time through the data communication using the network,
Constraints on unit tasks performed by the plurality of robots include location and time constraints,
The temporal constraints include start time, end time, and duration. delete The method of claim 1,
Wherein said temporal constraints can be set in whole or in part when planning a robot's task in a server. The method of claim 1,
The server is a multi-robot control system, characterized in that the real-time management in consideration of the position and movement of the robot for each unit time interval in the robot operation task for the occurrence of inconsistency and sudden situation for each unit task. The method of claim 4, wherein
The server is a multi-robot control system, characterized in that to add a waypoint / mission / mission in real time to the operation task of the robot. The method of claim 1,
When the server is inconsistent with the constraints on the performance of the task of the robot, the server may select any one of the following factors: determination, curtailment, cancellation, stand-by, pull, and extension. Multi-robot control system, characterized in that the control of the robot in real time by selecting one control signal. Planning a task based on a network configuration using N unit tasks for a plurality of robots;
Assigning tasks to the plurality of robots;
Managing the tasks of each robot manually or automatically in order to resolve inconsistencies between tasks performed by a plurality of robots in preparation for the planned task;
Constraints on unit missions planned in the mission planning process include location and time constraints,
The temporal constraint includes a start time, end time and duration. The method of claim 7, wherein
The mission planning step,
Constructing a mission planning network using a starting point, a goal point, a waypoint, a mission point and a mission ship for the planning of the task;
The method for controlling a multi-robot, characterized in that it further comprises a method for setting the time constraints (start time, end time, duration) in whole or in part by unit mission. The method of claim 7, wherein
The process of assigning tasks to multiple robots is
A multi-robot control method comprising allocating an optimal task in consideration of dependency, robot collaboration, and task load. 10. The method of claim 9,
The process of allocating a task in consideration of the above-mentioned relationship,
in the presence of a relationship of _{'m i>} m _{i between'} m _i and m _i,
t ^e _i ≤ t ^s _i Where m _i is the i-th mission, m _{i '} is the successor of the i-th unit, t ^e _i is the end of m _i , and t ^s _i' is the start of m _{i '} Multi-robot control method, characterized in that assigning a mission as possible. 10. The method of claim 9,
The process of allocating a task in consideration of the robot collaboration is divided into mandatory collaboration and autonomous collaboration according to a given condition. The method of claim 11,
The forced collaboration t ^s _{i ^'s} ≤t ≤t _i ^e _i', or satisfies the condition of t ^s _{i '} ≤ t ^e _i ≤ t ^e _{i' and} the condition of X _ij + X _{i'j '} ≥ 2,
Here, a start time of t ^s _i is a unit of assignment m _i, t ^s _i and the end time of a start time of _"the unit mission m _i of the next duty m _{i ',} t ^e _i is a unit of assignment m _i, t ^e _i is the end time of the _"assignment unit is m _i m _i of the subsequent _task", when X _ij m _i is a unit of assignment is assigned to robot j is 1 or 0, and the variable, X _{i'j 'mission} is m _{i "is} When assigned to the robot j 'is a multi-robot control method characterized in that the variable is 1 or 0.

Images