Search

KR-20260062900-A - MOBILE ROBOT SYSTEM BASED ON NON-MATRIX TYPE WORKING ENVIRONMENTAL MOVING AND METHOD FOR CONTROLLING THE SAME

KR20260062900AKR 20260062900 AKR20260062900 AKR 20260062900AKR-20260062900-A

Abstract

The present invention provides a mobile robot system (1) and a control method thereof, comprising a plurality of mobile robot units (20) capable of driving in an irregular work environment defined by an irregular grid of nodes and edges, and a mobile robot server (10) that receives planning information including at least a destination and a destination, checks whether there is a collision between the plurality of mobile robot units (20), transmits and provides a driving path to the mobile robot units, receives feedback of mobile robot status information transmitted from the mobile robot units (20), and searches, updates, and confirms the driving path of the mobile robot units (20); wherein, if the driving path includes rotational driving, the system generates neighbor nodes of the target node within an upper bound of a preset radius from the target node prior to rotational driving, and forms a curve path that enters onto the edge of the neighbor nodes after rotational driving to search and generate the driving path, and calculates a minimum cost driving path by differentiating the cost of the driving path according to whether the node is movable during the driving path search.

Inventors

  • 정민국
  • 김재권
  • 강소라
  • 박용희
  • 유재성
  • 최명서

Assignees

  • 주식회사 나비프라

Dates

Publication Date
20260507
Application Date
20251029
Priority Date
20241029

Claims (20)

  1. A plurality of mobile robot units (20) capable of driving in an irregular work environment that can be defined by an irregular grid of nodes and edges, and The system includes a mobile robot server (10) that receives planning information including at least a destination and a destination, checks for collision between the plurality of mobile robot units (20), transmits and provides a driving path to the mobile robot units, and receives feedback input of mobile robot status information transmitted from the mobile robot units (20) to search, update, and confirm the driving path of the mobile robot units (20). A mobile robot system (1) that, when the driving path includes rotational driving, generates neighbor nodes of the target node within an upper bound of a preset radius from the target node prior to rotational driving, forms a curve path that enters onto the edge of the neighbor node after rotational driving to search for the driving path, and calculates a minimum cost driving path by differentiating the cost of the driving path according to whether the node is movable during the search for the driving path.
  2. In Article 1, The above neighboring node is: Direct neighbor nodes directly connected to the above target node via an edge, and It includes an indirect neighbor node that is indirectly connected to the target node as a node within the upper bound, The above mobile robot server (10) is: A mobile robot system (1) characterized by including a robot path finding module (40) that receives the above robot status information and calculates the driving path of the above mobile robot unit (20).
  3. In Paragraph 2, The above robot path finding module (40) is: An upper level finding unit (410) that compares driving paths between the mobile robot units (20) to check for collisions and generates a constraint table including an expected node and an expected time at which a collision is expected to occur between the mobile robot units (20), and A mobile robot system (1) characterized by including a lower level finding unit (420) that searches for and regenerates a driving path of the mobile robot unit (20) using the constraint table generated in the upper level finding unit (410), wherein if the driving path includes rotational driving, the lower level finding unit (420) generates the driving path by forming one or more sampling curve paths that connect one or more sampling nodes formed on the edge between the target node before rotational driving and the direct neighbor node and the indirect neighbor node after rotational driving.
  4. In Paragraph 3, A mobile robot system (1) characterized in that the above sampling curve path includes a Bézier curve.
  5. In Paragraph 3, The above lower level finding section (420) is: A lower level state information forming unit (4210) that calculates lower level state information including the constraint table generated in the upper level finding unit (410), the current position information and time information of the mobile robot unit (20), and path attribute information for nodes and edges constituting the driving path of the mobile robot unit (20) including the sampling curve path when the driving path includes rotational driving, and calculates the corresponding cost function to re-search and generate a driving path, and A lower level termination determination unit (4220) that checks whether the driving path to the destination of the mobile robot unit (20) is completed, and A mobile robot system (1) characterized by including a node edge attribute generation unit (4230) that generates attribute information for nodes and edges forming the driving path by considering whether entry is possible from a target node constituting the driving path of the lower level state information to a neighbor node adjacent to the target node and the driving type, and transmits the information to the lower level state information forming unit (4210).
  6. In Paragraph 5, The lower level status information forming unit (4210) above is: A lower level state information generation unit (4211) that updates driving path information by calculating a cost function of a driving path for a target mobile robot unit among the mobile robot units (20) using the constraint table, node attribute information, edge attribute information, and the sampling curve path generated in the upper level finding unit (410) and the driving path when the driving path includes rotational driving, and A mobile robot system (1) characterized by including a lower level state information selection unit (4213) for selecting lower level state information for a driving path generated by the lower level state information generation unit (4211).
  7. In Paragraph 5, The above node edge attribute generation unit (4230) is: A node existence confirmation unit (4231) confirming the existence of neighbor nodes directly or indirectly adjacent to the target node from the target node constituting the driving path, and A movable node verification unit (4233) that checks whether movement is possible for a verified neighbor node, and A mobile robot system (1) characterized by including a node attribute assigning unit (4235) that assigns a node attribute according to the driving type of the mobile robot unit (20) to the neighbor node when movement to the neighbor node is possible, using information on whether the edge section connecting the target node and the neighbor node is curved or rotation angle.
  8. In Article 7, The above node attribute assignment unit (4235) is: A mobile robot system (1) characterized by setting the target node to a straight driving node attribute according to the driving type among forward driving, backward driving, and meandering driving when the mobile robot unit (20) can drive straight from the target node to the neighbor node without stopping, and setting the target node and/or the neighbor node to one of the straight forward driving node attribute, straight backward driving node attribute, and straight meandering driving node attribute.
  9. In Article 7, The above node attribute assignment unit (4235) is: The above node attribute assignment unit (4235) is characterized by setting the target node as a curved driving node attribute according to the driving type among forward driving, backward driving, and meandering driving when the mobile robot unit (20) stops at the target node and can drive to the neighbor node, and setting the target node and/or the neighbor node as one of the curved forward driving node attribute, curved backward driving node attribute, and curved meandering driving node attribute.
  10. In Article 7, The above node edge attribute generation unit (4230) is: A mobile robot system (1) characterized by including an edge attribute assigning unit (4237) that assigns edge attributes to an edge section connecting the end node and the target node using the node attributes of the end node of the re-searched driving path and the target node.
  11. In Article 10, The lower level state information forming unit (4210) updates the driving path information by calculating a cost function of the driving path for a target mobile robot unit among the mobile robot units (20) using the constraint table, the node attribute information, and the edge attribute information generated by the upper level finding unit (410). The edge attribute assignment unit (4237) forms edge attributes by assigning a velocity profile for each node section within the driving path according to the driving node attributes, and The mobile robot system (1) is characterized in that the lower level state information forming unit (4210) calculates a cost function value through the node attributes and edge attributes of the driving path.
  12. A providing step (S1) of providing a mobile robot system (1) comprising a plurality of mobile robot units (20) capable of driving in an irregular work environment definable as an irregular grid of nodes and edges, and a mobile robot server (10) that searches, updates, and confirms the driving path of the mobile robot units (20); A preparation step (S2) in which planning information including destinations and destinations and mobile robot status information transmitted from multiple mobile robot units (20) are collected and prepared, and A path search step (S10) in which a driving path is predicted in the robot path finding module (40) of the mobile robot server (10) based on the planning information and the mobile robot status information, and a driving path that is confirmed to be non-collision-free is calculated by checking whether there is a collision between the mobile robot units (20), and The method includes a path driving execution step (S70) in which the mobile robot unit (20) drives according to the driving path information determined in the path search step (S10) and transmits the current robot status information of each driving path to the mobile robot server (10). A mobile robot system control method characterized in that, in the path search step (S10), if the driving path includes rotational driving, neighbor nodes of the target node within an upper bound of a preset radius from the target node prior to rotational driving are generated, and a curve path entering onto the edge of the neighbor node after rotational driving is formed to search and generate the driving path, and a minimum cost driving path is calculated by differentiating the cost of the driving path according to whether the node is movable during the driving path search.
  13. In Paragraph 12, The above neighboring node is: Direct neighbor nodes directly connected to the above target node via an edge, and It includes an indirect neighbor node that is indirectly connected to the target node as a node within the upper bound, The above path search step (S10) is: An upper level state information forming step (S30) in which upper level state information including a constraint table of collision spatiotemporal node information of spatiotemporal overlap nodes between the driving paths of the mobile robot unit (20) is formed using the above planning information and the above mobile robot state information, and A lower level search step (S40) in which lower level state information including driving path information of the mobile robot is searched and generated using the attributes of the target node on the upper level state information and the mobile robot state information, wherein if the driving path includes turning, one or more sampling curve paths are formed connecting one or more sampling nodes formed on the edge between the target node before the turning path and the direct neighbor node and the indirect neighbor node after the turning path, thereby generating the driving path; An upper level state information update step (S100) that updates the upper level state information using the driving path information of the mobile robot generated in the lower level search step (S40), and A mobile robot system control method characterized by including a collision detection step (S110) that checks the collision spatiotemporal node information of the spatiotemporal overlap nodes between the driving paths of the mobile robot unit (20) using the planning information and the mobile robot status information to update and calculate the constraint table.
  14. In Paragraph 13, The planning information confirmed in the planning information confirmation step (S20) includes node edge information having node location, node time, and node edge attributes that form a driving path, based on the driving path information of the mobile robot unit (20). The above lower level search step (S40) is: A lower level state information forming step (S41) in which a cost function value for each node section according to the node attribute constituting the driving path of the mobile robot unit (20) including the sampling curve path is calculated and the driving path is updated when the driving path includes rotational driving, and A lower level termination confirmation step (S43) in which it is confirmed whether the driving path update of the mobile robot unit (20) is completed, and A mobile robot system control method characterized by including a neighbor node verification step (S45) that checks whether a neighbor node exists and whether entry is possible on the rear side of the target node.
  15. In Paragraph 14, The above lower level state information forming step (S41) is: A lower level state information generation step (S411) in which, when the driving path includes rotational driving, a cost function value for each node section according to the node and edge attributes according to the node attributes constituting the driving path of the mobile robot unit (20) including the sampling curve path is calculated and lower level state information including the driving path is updated and calculated, and A mobile robot system control method characterized by including a lower level state information selection step (S413) for selecting a target node among the driving paths using the position information of the mobile robot unit (20).
  16. In Paragraph 15, The above lower level state information generation step (S411) is: At least a lower level pre-check step (S4111) in which graph information of the above driving path is confirmed, and If the above driving path includes turning, an upper bound verification step (S4113) in which the direct neighbor node and the indirect neighbor node within an upper bound of a preset radius from the target node prior to turning are identified, and A sampling node formation step (S4114) of forming one or more sampling nodes on the edge with the indirect neighbor node after rotational driving, and A driving path forming step (S4115) in which one or more driving paths are generated, including a target node and a sampling curve that can be formed with a curve among the sampling nodes, and A cost calculation step (S4117) in which a driving path cost for the above driving path is calculated, and A mobile robot system control method characterized by including an optimal path determination step (S4119) for calculating a driving path having the minimum value among the driving path costs as the minimum cost driving path.
  17. In Paragraph 16, If type information of the above mobile robot is included, The above lower level pre-check step (S4111) is: A type information verification step (S4111a) in which the type information of the mobile robot is verified, and A mobile robot system control method characterized by including a graph information verification step (S4111b) in which the above graph information is verified.
  18. In Paragraph 16, The lower level state information generation step (S411) performs cost calculation for a plurality of preset type information of the mobile robot, and A mobile robot system control method characterized by having an optimal type confirmation step (S4118) in which type information and a driving path are confirmed, between the cost calculation step (S4117) and the optimal path confirmation step (S4119), in which the minimum driving path cost is calculated.
  19. In Paragraph 14, The above neighbor node verification step (S45) is: A neighbor node check step (S451) for checking neighbor nodes at the rear end of the above target node, and A neighbor node existence determination step (S452) for determining whether a neighbor node identified in the neighbor node check step (S451) exists, and A mobile robot system control method characterized by including a neighbor node entry possibility determination step (S453) for determining whether entry into a neighbor node is possible when it is determined that a neighbor node of the target node exists in the neighbor node existence determination step (S452).
  20. In Paragraph 19, The above neighbor node verification step (S45) is: If it is determined in the neighbor node entry possibility determination step (S453) that entry to the neighbor node is possible, a turning driving determination step (S458) that determines whether turning driving is necessary between the target node and the neighbor node and the driving type, and A mobile robot system control method characterized by including a node attribute determination step (S459) in which the node attribute of the target node and the attribute of the neighbor node are determined according to the result of the determination in the turning driving determination step (S458).

Description

Mobile Robot System Based on Non-Matrix Type Working Environment Moving and Method for Controlling the Same The present invention relates to a mobile robot server, a mobile robot system, and a control method thereof, and more specifically, to a mobile robot system, a mobile robot, and a control method that calculates a driving path and provides integrated control for enabling efficient driving of a plurality of mobile robot units. Robots are expanding their range of applications in various industrial fields and daily life. From robot vacuum cleaners used in homes to manufacturing robots used in welding or press processes in automobile factories, as well as various robots used in logistics systems such as AGVs and AMRs, a wide variety of robots are being researched, produced, and used. In particular, the scope of utilization and application of transport and logistics robots that are guided or autonomously driven, such as AGVs (Automatic Guided Vehicles) or AMRs (Autonomous Mobile Robots), is increasing dramatically. That is, due to the development of drive technology, beyond the need for operation in clean facilities such as semiconductor factories, there is a trend toward expanded application in transportation service industries such as shopping malls and freight transport, as well as various logistics delivery industries operating fulfillment systems, and even in heavy industries that require the transport of large loads, such as heavy industry. In the case of conventional AGVs or AMRs, particularly AGVs, they drove by following a floor line as a guide line within the driving environment or by providing passive or active markers such as optical identification codes like QR codes or additional external sensors. However, when complex path formation is required in the driving environment or when multiple robots are operating, complex challenges such as path crossing or avoidance are involved. Furthermore, in cases where multiple conventional autonomous vehicles operate in intersecting directions, a method of simply assigning priorities among the robots was adopted to resolve deadlocks between the vehicles. However, this priority assignment method entailed problems, such as frequent waiting actions to avoid path overlap in specific areas, maximized costs due to excessive waiting times for specific autonomous vehicles based on priority, and difficulties in overall integrated control. In addition, in the case of conventional unmanned vehicles, even if the vehicles have the same specifications, the degree of fatigue accumulation varies due to repetitive use, making it difficult to accurately predict the path using only simple physical size information. That is, in the case of conventional unmanned vehicles, differences in driving speed and responsiveness occur even among vehicles of the same specifications depending on the degree of fatigue accumulation and aging, causing errors. These errors accumulate as driving time increases, making it difficult to accurately predict the location. This entails the disadvantage that when multiple unmanned vehicles are driving, disturbances caused by unexpected deadlocks at multiple points are maximized. Furthermore, in the case of conventional autonomous vehicles, while integrated control of multiple vehicles based on node occupancy status is suitable for simple, unidirectional environments, it entails a problem where efficiency plummets in complex driving environments, such as unstructured environments where it is difficult to form nodes in the form of a standardized grid map. Although the possibility of collisions was eliminated by providing real-time vehicle location information to enable position identification via communication between autonomous vehicles and path selection through evasive maneuvers to resolve this issue, it still entails difficulties. These include increased frequency of safety accidents caused by individualized driving in environments with a large number of vehicles, such as those requiring heavy and heavy vehicle transport where rapid evasive maneuvers are restricted, or a sharp decline in driving efficiency under overall integrated conditions. Furthermore, when multiple autonomous vehicles are in operation and the vehicles attempt to incorporate real-time information, the computational load required to predict each vehicle's driving path increases exponentially. This leads to increased computation time due to the computational load, and this increase in computation time also entails the problem of causing a discrepancy with the actual driving environment, resulting in meaningless computational weighting. FIG. 1 is a schematic overall configuration diagram of a mobile robot system according to an embodiment of the present invention. FIG. 2 is a schematic diagram of the configuration of a mobile robot server of a mobile robot system according to an embodiment of the present invention. FIGS. 3 to 7 are schematic diagrams of the detailed configuration of