CN115755786A

CN115755786A - Multi-AGV global scheduling method based on flow

Info

Publication number: CN115755786A
Application number: CN202211323177.6A
Authority: CN
Inventors: 任鹏举; 毛艺钧; 丁焱; 焦崇珊; 杨勐; 郑南宁
Original assignee: Xian Jiaotong University
Current assignee: Xian Jiaotong University
Priority date: 2022-10-26
Filing date: 2022-10-26
Publication date: 2023-03-07

Abstract

A flow-based multi-AGV global scheduling delivery system comprises a task allocation module, a global path planning module and a delivery control module; the task allocation module is used for allocating tasks and binding the tasks to an AGV to execute the tasks; the global path planning module is used for planning a global path to a target location for all AGVs with moving requirements by using a multi-AGV collision-free path planning method based on path expected flow; the traffic control module is used for monitoring the running state of each AGV in real time, and according to the real-time flow and the path attributes of the remaining paths, if the residual paths can continue to advance, the residual path sequences and the corresponding path attributes are sent to each AGV in a segmented mode to be executed, and if the residual paths cannot continue to advance, the AGV waits for commands in place or replans the paths by taking the current position as a starting point. The system can avoid the condition of multiple AGV paths conflict or congestion, and ensures the high-efficiency operation of the whole system.

Description

Multi-AGV global scheduling method based on flow

Technical Field

The disclosure belongs to the technical field of AGV intelligent control, and particularly relates to a flow-based multi-AGV global scheduling method.

Background

An Automated Guided Vehicle (AGV) is a transport device that can automatically perform material handling, and it is usually navigated using radio, camera, laser radar, or magnetic stripes, magnetic nails, and two-dimensional codes marked on the ground. Compared with other logistics transportation equipment, the AGV has the advantages of being strong in adaptability, high in automation degree, capable of saving labor cost, convenient to maintain and the like. With the gradual increase of labor cost and the increasingly varied production patterns, more and more enterprises adopt highly automated production systems, and the AGVs are important components of the automated production systems. Therefore, the design research of the AGV has important significance for improving the production efficiency and reducing the production cost of enterprises.

Compare with single AGV, the ability that many AGV systems carried the transport operation is stronger to in the face of complicated changeable factory environment, many AGV systems can the quick response external environment's change, and the task is accomplished to nimble efficient, consequently develops many AGV systems just in order to be necessary. Compared with a single AGV, the multi-AGV system needs to solve the problems of multi-machine cooperation, information interaction, conflict resolution and the like, so that the design of the multi-AGV system is more complex. At present, designing a multiple AGV system with high population, self-organization and self-adaptability becomes a research hotspot, mainly including the following research directions: the method comprises the following steps of research of an action analysis and control algorithm, research of an autonomous perception and networking algorithm, research of a multi-machine positioning algorithm, research of a multi-machine autonomous dynamic decision and path planning algorithm, research of a multi-machine formation combination solution, research of a bionic behavior simulation algorithm and the like.

The problem of scheduling multiple AGVs is also a key point of the multiple AGV system. The purpose of this problem is to plan an appropriate global path for each of a plurality of AGVs in the entire AGV system so that they can go from the start point to the end point and no collision occurs. Since there is a shared path between multiple AGVs, it is necessary to control the multiple AGVs in order to solve resource competition between the AGVs during traveling. At present, two methods for dispatching and controlling the AGV are provided, namely centralized control and distributed control. Centralized control means that a scheduling center performs unified scheduling planning on all the AGVs, and distributed control means that each AGV plans for itself independently. The method adopts a centralized control method, and has the advantages of simple structure, convenience in control, higher scheduling efficiency and the like.

On the centralized control, two main categories can be distinguished: the first is to resolve the problem into a single AGV global path plan, and then when the AGV walks, a plurality of AGV collide with each other by using certain traffic management rules or walking rules and the like; the second type is to consider when and where the AGVs will be, and directly plan a non-collision path with time constraints for each AGV, so that multiple AGVs will not collide under the condition that the AGVs travel strictly according to the path requirements. However, the two methods have their own drawbacks, for example, the first method cannot take into account the routes of other AGVs during planning, so that the planned route may cause congestion on a certain section of road or deadlock among AGVs, and it is difficult for traffic management rules or walking rules to resolve conflicts among AGVs; the second method is lack of flexibility, and is only suitable for being used in a scene without random interference, and in a scene with more random interference, the time constraint requirement of the path of the AGV in the traveling process may be difficult to meet.

Disclosure of Invention

In order to solve the technical problem, the disclosure discloses a flow-based multi-AGV global scheduling traffic control system, which comprises a task allocation module, a global path planning module and a traffic control module; wherein,

the task allocation module is used for allocating the tasks and binding the tasks to an AGV to execute the tasks;

the global path planning module is used for planning a global path to a target location for all AGVs with moving requirements by using a multi-AGV collision-free path planning method based on path expected flow;

the traffic control module is used for monitoring the running state of each AGV in real time, and according to the real-time flow and the path attributes of the remaining paths, if the residual paths can continue to advance, the residual path sequences and the corresponding path attributes are sent to each AGV in a segmented mode to be executed, and if the residual paths cannot continue to advance, the AGV waits for commands in place or replans the paths by taking the current position as a starting point.

By the technical scheme, the traffic management system is globally scheduled to interface service logic based on the multiple AGVs of the flow, the navigation tasks are received, one current idle AGV is selected, and the tasks are distributed to the AGVs; and after the task allocation is finished, planning the path for each running AGV according to the expected flow of each path. Because the application range of the AGV system comprises a wide path capable of accommodating a plurality of AGVs to run side by side, the scheme provides a multi-AGV collision-free path planning method based on the path expected flow. And after the global path is obtained, monitoring the running state of each AGV in real time, issuing a residual path sequence and corresponding path attributes to each AGV for execution in a segmented manner if the AGV can continue to advance according to the real-time flow and the path attributes of the residual path, and enabling the AGV to wait in situ or re-plan the path if the residual path sequence and the corresponding path attributes cannot be issued, so that the condition of multiple AGV paths conflict or congestion is avoided, and the efficient running of the whole system is guaranteed.

Drawings

FIG. 1 is a block diagram flow diagram of a traffic-based multiple AGV global scheduling scheme provided in one embodiment of the present disclosure;

FIG. 2 is a block diagram illustrating a flow chart of a method for multiple AGV collision-free path planning based on expected path flows according to an embodiment of the present disclosure;

fig. 3 is a block flow diagram of a method for selecting a release path provided in an embodiment of the present disclosure.

Detailed Description

In order to make those skilled in the art understand the technical solutions disclosed in the present disclosure, the technical solutions of various embodiments will be described below with reference to the embodiments and the accompanying fig. 1 to 3, where the described embodiments are some embodiments of the present disclosure, but not all embodiments. The terms "first," "second," and the like as used in this disclosure are used for distinguishing between different objects and not for describing a particular order. Furthermore, "include" and "have," as well as any variations thereof, are intended to cover and not to exclude inclusions. For example, a process, method, system, or article or apparatus that comprises a list of steps or elements is not limited to only those steps or elements but may alternatively include other steps or elements not expressly listed or inherent to such process, method, system, article, or apparatus.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the disclosure. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It will be appreciated by those skilled in the art that the embodiments described herein may be combined with other embodiments.

Referring to fig. 1, in one embodiment, the present disclosure discloses a traffic management system for multiple AGVs global scheduling based on traffic, which includes a task allocation module, a global path planning module, and a traffic management control module; wherein,

For the purposes of this embodiment, the system is primarily directed to distributing tasks and binding them to an AGV to perform the task. Meanwhile, for all the AGVs with the moving requirements, a multi-AGV collision-free path planning method based on the expected path flow is used for planning a global path to a target location, the running state of each AGV is monitored in real time through a traffic control module, and according to the real-time flow and the path attributes of the remaining paths, a remaining path sequence and corresponding path attributes are selected to be issued in a segmented mode, or an in-place waiting command is selected to be issued, or the path is re-planned by taking the current position as a starting point, so that the condition of multi-AGV path collision or congestion is avoided, and the efficient operation of the whole system is guaranteed.

As shown in fig. 1, the communication protocol for uploading and sending of the system is specified. The communication protocol for the AGV to upload the system is V2S _ protocol, and mainly comprises contents such as an online request, position initialization, state information, a response instruction and the like; the communication protocol for issuing the AGV by the system is S2V _ protocol, and mainly comprises contents such as initialization response feedback, online response feedback, path types, path numbers, task instructions and the like.

In another embodiment, in the task allocation module, it is necessary to continuously check whether there is a new task requirement; when finding that a new task is needed, adding the new task into an unallocated task list; when scene information is updated every time, whether idle AGV and unallocated tasks exist currently needs to be checked; if not, continuously checking whether a new task exists; if yes, distributing the unallocated task with the task location closest to the current position of the AGV to the newly uploaded AGV, and deleting the distributed task in an unallocated task list.

For this embodiment, in the task assignment module, the module needs to constantly check whether new navigation task needs are sent to the system. When a new task is found to be needed, the new task needs to be added into the unallocated task list. At each scene information update, it is necessary to check whether there is currently an empty AGV, i.e. one that is not assigned a task. If no idle AGV exists, continuously checking whether a new task exists; if yes, selecting the task with the task place closest to the current position of the AGV from the unallocated task list for the idle AGV which uploads the information to the scheduling module, allocating the task to the AGV, and deleting the allocated task from the unallocated task list.

In another embodiment, the method for planning a collision-free path of multiple AGVs based on expected path traffic specifically includes:

on a topological map obtained by modeling a real environment, based on a classic path planning algorithm A-algorithm, the attributes of the paths and the expected flow of the paths are considered in the calculation process of the cost function, and a proper global path is planned, so that the time consumed by AGV driving is minimized, and the probability of congestion of multiple AGVs is reduced.

For the embodiment, in order to complete global path planning for the AGVs, a brand-new multi-AGV collision-free path planning method based on the expected path traffic is designed. According to the method, on a topological map obtained by modeling a real environment, a classical path planning algorithm A-x algorithm is taken as a basis, the attributes and expected flow of paths are considered in the calculation process of a cost function, a proper global path is planned, the time consumed by AGV driving is minimized, and the possibility of congestion of multiple AGVs is reduced.

Different from the common A-star algorithm, a new cost G value calculation method based on the path expected flow is designed. When the node is expanded, if the current node is N and the node to be expanded is A, the path from N to A can pass and is not blocked, the cost G of the node A _A The value can be expressed as:

wherein G is _N Is the cost G value, d of the current node N _NA Is the length of the path from node N to node A, r _NA Minimum camber value of deflection, e, is required to take the path from node N to node A _NA For the expected traffic of node N to node A path, m _NA The maximum flow of the path from the node N to the node A is obtained from the path attributes, and mortar, beta and gamma are constants for adjusting the preference of the generated path and can be set according to actual conditions. The cost G value of the initial node is 0.

H of node A _A The value defines the straight to the target nodeLine distance, final F _A The value can be represented by F _A ＝G _A +H _A And (4) calculating.

The specific flow of the multiple AGV collision-free path planning method based on the expected flow of the path is shown in fig. 2. The locations of the start and end points of the AGV, as well as the existing expected traffic of the current topological map, need to be given first. The OPEN table and CLOSE table are defined, initialized to null, and added to the OPEN table after calculating their F values for the starting point. Using one cycle, it is determined whether it is empty for the OPEN table each time. If the path is empty, the path solving is failed; if not, selecting the node N with the minimum F value in the OPEN table at the moment, and putting the node N into the CLOSE table. Judging whether the node N is an end point, if so, indicating that the solution is successful, reversely searching a preorder node of the node N, generating an optimal path, updating the expected flow and outputting the path; if not, acquiring all adjacent node sets Adj of the node N, sequentially taking out each adjacent node for expansion judgment until the sets Adj are empty, entering the next cycle at the moment, and judging whether the OPEN table is empty or not.

When expanding the adjacent node A, firstly judging whether the AGV can go from the node N to the node A, if not, ending the expansion judgment of the node; if so, the value of G through node N to node A is determined based on the expected flow. Searching whether the node A is in an OPEN table, if so, judging whether a new G value is smaller than an old G value, if so, deleting the node A in the OPEN table, calculating an H value and an F value of the node A, adding the node A into the OPEN table again, finishing the expansion judgment, and if not, finishing the expansion judgment on the node; if the node A is not in the OPEN table, it needs to be determined whether it is in the CLOSE table. If the new G value is smaller than the old G value, judging whether the new G value is smaller than the old G value and carrying out the same subsequent operation; and if not, calculating the H value and the F value of the node A, adding the H value and the F value into the OPEN table, and finishing the expansion judgment.

In fig. 2, it should be noted that after taking out a node a from all the neighboring node sets Adj of the current node N, it needs to be determined whether the AGV can go from the node N to the node a. Because the real environment may contain a path which can only pass in one way, and the path in the topograph is non-directional, when the node needs to be expanded, whether the path from N to A can pass or not needs to be checked according to the real environment. In addition to this, it is also necessary to check whether the path from N to a is blocked. If the planning of the path fails, the scheduling is directly finished, and warning information is displayed.

In another embodiment, the path expected traffic is updated in two cases: the first case is when the real-time scene information is updated; the second case is after the optimal path is generated.

For this embodiment, the path expectation traffic is updated in two cases. The first situation is that when the real-time scene information is updated, if the AGV finishes the previous path, the expected flow of the previous path is reduced by 1, which indicates that the AGV has passed through the path. The second case is that after the optimal path is generated, the expected traffic for each segment of the global path needs to be increased by 1, indicating that the AGV is going to traverse these paths.

In another embodiment, in the traffic control module, the remaining path of each AGV is updated first; then judging whether a residual path exists or not; if not, the AGV is considered to have arrived at the task end point, the task is completed, the AGV is set to be in an idle state, and the scheduling is finished; if yes, checking whether the AGV needs to be released; if not, finishing the scheduling, and waiting for next scheduling; if yes, checking whether the next section of path in front of the AGV is blocked, if the path in front of the AGV is blocked, taking the end point of the path where the AGV is currently located as a starting point and taking a target point as an end point, eliminating the influence of the original remaining path on the expected flow of the corresponding path, planning the global path again, and executing the subsequent flow; if the path is not blocked, selecting a path which can be released from the remaining paths; judging whether a releasable path exists, if not, issuing a waiting command, and finishing the scheduling; if yes, updating the real-time flow of the releasable path, simultaneously issuing a releasable path sequence and a corresponding path attribute, and finishing the scheduling.

In this embodiment, each route has its maximum flow limit, so as to avoid danger or congestion caused by the fact that the flow of AGVs on a certain route exceeds its maximum flow limit, the system implements traffic management of multiple AGVs, i.e. segmented issuing and releasing of the remaining global routes of multiple AGVs, by using the traffic management control module, and the specific flow is as shown in fig. 1.

After each time of updating the real-time scene information, the traffic control module needs to update the real-time traffic of each path in the topological map. When the AGV first travels to or from the end of a certain path along the path, the real-time traffic of the path is reduced by 1. After the system determines the pass path for the AGV, the real-time traffic of all the paths that can be passed is increased by 1.

In addition, the system requires a schedule for each AGV. The first is to update the remaining path for each AGV. After the remnant path is updated, judging whether the remnant path exists: if not, the AGV is considered to have already arrived at the task end point, the task is completed, the AGV is set to be in an idle state, and the scheduling is finished; if so, it is first checked whether the AGV needs to be released. When it is determined that release is required, first, it is checked whether the next path in front of the AGV is blocked. If the front path is blocked, the terminal point of the current path of the AGV is taken as the starting point, the target point is taken as the terminal point, the influence of the original remaining path on the expected flow of the corresponding path is eliminated, the global path planning is carried out again, and the subsequent process is executed; if the path is not blocked, a path which can be released is selected from the rest paths. If no path is available, issuing an in-place waiting command to the AGV; if the releasable path exists, updating the real-time flow of the releasable path, simultaneously issuing a corresponding path sequence and attribute, and finishing the scheduling.

In another embodiment, a specific method for selecting a path that can be released from the remaining paths is as follows: initializing a releasable list; taking a first section of the remaining paths as a candidate path; judging whether the real-time flow of the candidate path is smaller than the maximum flow, if so, adding the candidate path into a releasable path list, and deleting the candidate path from the rest paths; if not, outputting a releasable path list, and finishing the selection; judging whether the total length of the paths in the releasable path list is greater than a threshold value, if so, outputting the releasable path list, and ending the selection; if not, returning to take the first section of the remaining paths as a candidate path, and continuing to judge until the selection is finished.

For the embodiment, the specific flow method is shown in fig. 3, and the size of the threshold value can be set according to the actual situation.

In another embodiment, the maximum flow rate is the number of AGVs traveling in the same direction that can pass side by side at the narrowest point of the entire path; the real-time traffic is how many AGVs are currently or will pass through the path for all the pass paths.

For this embodiment, in order to meet the worst-case requirement that multiple AGVs can still theoretically safely pass through one segment of the path, the maximum flow rate of one segment of the path is defined as the maximum number of AGVs that the full Cheng Bingpai can safely pass through the path, i.e., the number of AGVs that can be accommodated side-by-side at the narrowest point of the path, i.e., the maximum flow rate. Thus, the expected traffic is how many AGVs are currently or will pass through the path for the global path; real-time traffic is how many AGVs are currently or will pass through the path for all the cleared paths.

In another embodiment, the specific steps for updating the remaining path of each AGV are: for each AGV, the current position of the AGV is located, so that the current path can be obtained, all paths before the path are deleted from the rest global paths, and if the current AGV is located at the end point of the path, the path is required to be deleted together.

For this embodiment, for each AGV, by locating the current position, it can obtain the current path, and delete all paths before the current path from the remaining global paths, and if the current AGV is located at the end of the current path, it needs to delete the current path together. After global path planning, the remaining paths also need to be updated. Specifically, all existing remnant paths are deleted, and a new global path is set as a remnant path.

In another embodiment, the specific criteria for checking whether the AGV needs to be released are: if the route is not issued to the AGV, or the distance between the AGV and the end point of the last section of the route issued last time is smaller than a set threshold value, the route is judged to need to be released, otherwise the route is not released, wherein the set distance threshold value is the route attribute of each section of the route.

For the embodiment, if the route has not been issued to the AGV, or the distance between the AGV and the end point of the last section of the route issued last time is smaller than the set threshold, it is determined that release is required, and the following process is performed; otherwise, the subsequent remaining path does not need to be issued, that is, the remaining path is not released, and the scheduling is finished, and the next scheduling of the AGV is waited. The set distance threshold is the path attribute of each path and needs to be given in the modeling of the topological map.

In another embodiment, the blocked paths refer to those paths that are blocked when the paths are completely impassable due to some obstacle or AGV and cannot resume traffic through the scheduling of the scheduling system itself.

Although the embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the above-described embodiments and application fields, and the above-described embodiments are illustrative, instructive, and not restrictive. Those skilled in the art, having the benefit of this disclosure, may effect numerous modifications thereto without departing from the scope of the invention as defined by the appended claims.

Claims

1. A flow-based multi-AGV global scheduling delivery system comprises a task allocation module, a global path planning module and a delivery control module; wherein,

the global path planning module is used for planning a global path to a target location for all AGVs with movement requirements by using a multi-AGV collision-free path planning method based on path expected flow;

2. The system of claim 1, preferably, in the task allocation module, it is required to continuously check whether there is a new task demand; when finding that a new task is needed, adding the new task into an unallocated task list; when scene information is updated every time, whether idle AGV and unallocated tasks exist currently needs to be checked; if not, continuously checking whether a new task exists; if yes, distributing the unallocated task with the task location closest to the current position of the AGV to the newly uploaded AGV, and deleting the distributed task in an unallocated task list.

3. The system according to claim 1, wherein said multiple AGV collision-free path planning method based on expected path traffic is specifically:

on a topological map obtained by modeling a real environment, based on a classic path planning algorithm A algorithm, the attribute of a path and the expected flow of the path are considered in the calculation process of a cost function, and a proper global path is planned, so that the time consumed by AGV driving is minimized, and the possibility of congestion of multiple AGVs is reduced.

4. The system of claim 3, wherein the path expected traffic is updated in two cases: the first case is when the real-time scene information is updated; the second case is after the optimal path is generated.

5. The system of claim 1, wherein the traffic control module first updates the remaining path of each AGV; then judging whether a residual path exists or not; if not, the AGV is considered to have arrived at the task end point, the task is completed, the AGV is set to be in an idle state, and the scheduling is finished; if yes, checking whether the AGV needs to be released; if not, finishing the scheduling, and waiting for next scheduling; if yes, checking whether the next section of path in front of the AGV is blocked, if the path in front of the AGV is blocked, taking the end point of the path where the AGV is currently located as a starting point and taking a target point as an end point, eliminating the influence of the original remaining path on the expected flow of the corresponding path, planning the global path again, and executing the subsequent flow; if the path is not blocked, selecting a path which can be released from the remaining paths; judging whether a releasable path exists, if not, issuing a waiting command, and finishing the scheduling; if yes, updating the real-time flow of the releasable path, simultaneously issuing a releasable path sequence and a corresponding path attribute, and finishing the scheduling.

6. The system of claim 5, wherein the specific method for selecting the traversable paths from the remaining paths is as follows: initializing a releasable list; taking a first section of the remaining paths as a candidate path; judging whether the real-time flow of the candidate path is smaller than the maximum flow, if so, adding the candidate path into a releasable path list, and deleting the candidate path from the rest paths; if not, outputting a releasable path list, and finishing the selection; judging whether the total length of the paths in the releasable path list is greater than a threshold value, if so, outputting the releasable path list, and ending the selection; if not, returning to take the first section of the remaining path as a candidate path, and continuing to judge until the selection is finished.

7. The system of claim 6, wherein said maximum flow rate is the number of AGVs that can pass side by side traveling in the same direction at the narrowest point of the overall path; the real-time traffic is how many AGVs are currently or will pass through the path for all the pass paths.

8. The system of claim 5, wherein the step of updating the remaining path of each AGV comprises: for each AGV, the current position of the AGV is located, so that the current path can be obtained, all paths before the path are deleted from the rest global paths, and if the current AGV is located at the end point of the path, the path is required to be deleted together.

9. The system of claim 5, wherein the specific criteria for checking whether the AGV needs to be released are: if the route is not issued to the AGV, or the distance between the AGV and the end point of the last section of the route issued last time is smaller than a set threshold value, the fact that the route needs to be released is judged, otherwise the route is not released, wherein the set threshold value is the route attribute of each section of the route.

10. The system of claim 5, wherein the blocked paths are specifically blocked when the paths are completely impassable due to some obstacle or AGV and cannot resume passing through the scheduling of the scheduling system itself.