A single lane scene based on turn-back avoidance multi-agv unlocking method

Abstract

The application discloses a kind of single lane scene under the method for unlocking multiple AGVs based on turn-back avoidance, specific steps include being formed directed graph according to AGV current position point and next position point, the execution route of AGV is distributed;Determine whether directed graph is looped, check whether there is deadlock between AGV;If there is deadlock, with the number of steps of retreating and the number of nodes allowed to avoid as comparison condition, select the most suitable AGV to actively avoid;According to the definition of allowed avoidance point, screen out avoidance point set P;Traversal avoidance point set P, it is as intermediate node, respectively plan two routes;The priority of AGV that actively avoids is less than AGV that is allowed to pass through in priority;By setting trigger, control the waiting and wake-up of AGV that actively avoids;Through the judgment condition set, select the best AGV to actively avoid, re-plan route, find out AGV that is allowed to pass through in priority, set trigger to control the waiting and wake-up of AGV that actively avoids, thereby solve the problem of single lane AGV deadlock.

Description

Technical Field

[0001] This invention relates to the technical field of industrial robots, and in particular to a method for unlocking multiple AGVs based on turning back and avoiding obstacles in a one-way street scenario. Background Technology

[0002] With the rapid development of information technology, the application of computers has gradually expanded from the internet industry to other fields such as industrial manufacturing. It is foreseeable that the application of computer technology in industrial manufacturing will become increasingly widespread, driving China's manufacturing industry towards digitalization and intelligence. Currently, most small and medium-sized factories use assembly line production, which has significantly improved production efficiency, but goods still require manual handling, resulting in a low level of automation. However, with the continuous development of global manufacturing and industrial automation, Automated Guided Vehicles (AGVs), as a new type of automated logistics and transportation equipment, are receiving increasing attention. Research on AGV scheduling aims to improve the automation level of factory logistics and manufacturing processes, optimize production processes, and improve production efficiency and quality.

[0003] AGV path planning refers to finding an optimal or efficient path for an AGV in a given environment, enabling it to reach a target location or complete a specific task. Path planning typically needs to consider various factors, such as the starting point, ending point, obstacles, road conditions, traffic rules, and cost functions. When multiple AGVs are searching for a path, the issue of AGV conflict needs to be considered. Using resource allocation scheduling strategies can ensure that AGVs do not conflict head-on or at nodes. However, for one-way streets, a single node conflict can extend to a conflict along a route. When two AGVs are traveling towards each other on a one-way street, a strategy is needed to adjust their priority and efficiency, ensuring that they can autonomously avoid collisions by replanning their offline routes when a deadlock occurs or is imminent.

[0004] This invention aims to solve the deadlock problem of multiple AGVs in the specific scenario of one-way streets by combining a resource allocation strategy with a turnaround and avoidance scheme. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0006] In view of the deadlock problem of multiple AGVs on existing one-way streets, the present invention is proposed.

[0007] Therefore, the purpose of this invention is to provide a multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario. Its purpose is to find an optimal or effective path for multiple AGVs in a given one-way street scenario so that they can reach the target location or complete a specific task, thereby solving the deadlock problem.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution, which includes the following steps: constructing a directed graph based on the current position and the next position of the AGV, and allocating a route for the AGV to execute;

[0009] To determine if a directed graph is cyclic, check for deadlocks between AGVs and denote the deadlocked AGV as V. i ;

[0010] If a deadlock occurs, the most suitable AGV is selected for active avoidance by comparing the number of backoff steps and the number of nodes allowed to be avoided. The AGV that actively avoids the deadlock is denoted as V. x ;

[0011] Based on the definition of permissible avoidance points, a set P of avoidance points is selected;

[0012] Iterate through the set of avoidance points P, using them as intermediate nodes, and plan the two routes respectively;

[0013] AGVs that actively avoid obstacles have lower priority than AGVs that have priority to pass.

[0014] By setting triggers, the waiting and waking of the AGV that actively avoids obstacles can be controlled.

[0015] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in the one-way scenario described in this invention, the number of AGVs includes at least two AGVs working simultaneously; the directed graph is based on a map model using the one-way as the main road to connect the production area and the shelf area; the one-way is provided with associated workstations, several leaf nodes, avoidance points and standby points.

[0016] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario described in this invention, wherein: the length of any path on the one-way street is sufficient to ensure that two AGVs do not collide when located at opposite ends; and on any path of the one-way street, when an AGV turns or rotates in place, it will not affect other AGVs.

[0017] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario described in this invention, the following is defined: the number of retreat steps refers to the number of steps required for the AGV to reverse-search to find the avoidance point along the list of completed road segments in the current route; the allowed avoidance point refers to the scheduling node where the AGV is allowed to avoid; if the number of retreat steps is small, the AGV with the fewest retreat steps is selected for active avoidance; if the number of retreat steps is the same, the AGV with the fewest allowed avoidance nodes is selected.

[0018] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in the one-way scenario described in this invention, the selection criteria for avoidance points include: the number of successor nodes is greater than 0; it is not occupied by an idle AGV; it is not associated with any workstation; and it is not on the subsequent route of the interlocked AGV.

[0019] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario described in this invention, wherein: from V x Select the allowed avoidance points from the direct successor nodes of the current node;

[0020] If the set of avoidance points P is empty, then it indicates that V x Unable to avoid directly, traverse V i If an unfinished section of the route is found to allow a detour, it is added to P.

[0021] If the set of avoidance points P is still empty, then traverse the standby points defined in the map and add all nodes that have not yet been occupied by the AGV to the set P.

[0022] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and obstacle avoidance in a one-way street scenario described in this invention, wherein: the two routes include from V x Given two routes, S1 from the current location to the avoidance point and S2 from the avoidance point to the destination node, if both routes exist, select the avoidance point with the lowest total cost between S1 and S2.

[0023] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario described in this invention, wherein: the priority of the AGV actively avoiding the obstacle is lower than that of the AGV with priority passage, including when V x When moving to an avoidance point, if its priority is higher, it will immediately acquire the resources of the original conflict point or conflict segment, which will lead to another deadlock. Therefore, V x To obtain conflict resources, all AGVs with priority passage must have released the conflict resources.

[0024] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in the one-way street scenario described in this invention, the trigger is set to ensure that the interlocked AGV can quickly wake up the avoidance AGV after leaving the conflict section. The basic working mode of the trigger includes blocking the avoidance AGV when it reaches the avoidance point, and waking up the active avoidance AGV when the priority AGV leaves the conflict section.

[0025] As a preferred embodiment of the multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario described in this invention, wherein: the triggers determine whether to allow the allocation of conflicting resources to V. x Have conflicting resources been released by the AGVs with priority passage?

[0026] The beneficial effects of this invention are as follows: In a one-way street scenario, when an AGV is deadlocked, the optimal AGV is selected for active avoidance based on the set judgment conditions. Then, the route is planned through the selected avoidance points to find the AGV with priority passage. Finally, by setting a trigger, the waiting and waking of the active avoidance AGV is controlled, thereby solving the deadlock problem. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0028] Figure 1 This is a flowchart of the reversal and avoidance strategy for the multi-AGV unlocking method based on reversal and avoidance in a one-way street scenario according to the present invention.

[0029] Figure 2 This is a single-lane example and execution route diagram of the multi-AGV unlocking method based on turnaround and avoidance in a single-lane scenario according to the present invention.

[0030] Figure 3 This is a schematic diagram of the AGV task execution of the multi-AGV unlocking method based on turning back and avoiding obstacles in a one-way street scenario according to the present invention.

[0031] Figure 4 This is a schematic diagram illustrating the deadlock of AGVs in the multi-AGV unlocking method based on turnaround and avoidance in a one-way street scenario according to the present invention.

[0032] Figure 5 This is a schematic diagram of the turning-back and avoidance process in the multi-AGV unlocking method based on turning-back and avoidance in a one-way street scenario according to the present invention.

[0033] Figure 6This is a schematic diagram illustrating the AGV deadlock release method based on turnaround and avoidance in a one-way street scenario according to the present invention. Detailed Implementation

[0034] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0035] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0036] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0037] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0038] Example 1

[0039] Reference Figure 1 This is the first embodiment of the present invention, which provides a method for unlocking multiple AGVs based on turnaround and obstacle avoidance in a one-way street scenario. The specific steps of this method include:

[0040] S1. Based on the directed graph formed by the current position and the next position of the AGV, allocate the execution route of the AGV.

[0041] The number of AGVs includes at least two AGVs working simultaneously; the directed graph is based on a map model and uses one-way streets as the main roads to connect the production area and the shelving area; the one-way streets are equipped with associated workstations, several leaf nodes, avoidance points and standby points.

[0042] The length of any path on a one-way street is sufficient to ensure that two AGVs do not collide when they are at opposite ends; on any path of the one-way street, when an AGV turns or rotates in place, it will not affect other AGVs.

[0043] S2, determine if the directed graph is cyclic, check for deadlocks between AGVs, and denote the deadlocked AGV as V.i .

[0044] When multiple AGVs begin executing the planned route, a depth-first search is used to check if the directed graph is cyclic.

[0045] The conflict between AGVs is handled based on a resource allocation strategy. AGVs are only allowed to move when the scheduling system allocates resources to them. When no further resources are allocated, they need to wait in place.

[0046] S3. If a deadlock exists, the most suitable AGV is selected for active avoidance based on the comparison between the number of backoff steps and the number of nodes allowed to be avoided. The AGV that actively avoids the deadlock is denoted as V. x .

[0047] The number of retreat steps refers to the number of steps required for the AGV to search backwards along the list of completed road segments in the current route to find the avoidance point; the allowed avoidance point refers to the scheduling node where the AGV is allowed to avoid; if the number of retreat steps is less, the AGV with fewer retreat steps is selected to actively avoid; if the number of retreat steps is the same, the AGV with fewer allowed avoidance nodes is selected.

[0048] The conditions for a successor node to serve as an avoidance point are as follows: the number of successor nodes is greater than 0; it is not occupied by an idle AGV; it is not associated with any workstation; and it is not on the subsequent route of an interlocked AGV.

[0049] S4. According to the definition of allowed avoidance points, select the set of avoidance points P.

[0050] From V x Select the permissible avoidance points from the direct successor nodes of the current node.

[0051] If the set of avoidance points P is empty, then it indicates that V x Unable to avoid directly, traverse V i For the unfinished part of the route, if an allowable avoidance point is found, it is added to set P; if the avoidance point set P is still empty, the standby points defined in the map are traversed and all nodes that have not yet been occupied by the AGV are added to set P.

[0052] S5, traverse the set of avoidance points P, use them as intermediate nodes, and plan two routes respectively.

[0053] The two routes include from V x Given two routes, S1 from the current location to the avoidance point and S2 from the avoidance point to the destination node, if both routes exist, select the avoidance point with the lowest total cost between S1 and S2.

[0054] S6, the priority of AGVs that actively avoid obstacles is lower than that of AGVs that have priority to pass.

[0055] AGVs that actively avoid obstacles have a lower priority than AGVs that have priority to pass, including when V x When moving to an avoidance point, if its priority is higher, it will immediately acquire the resources of the original conflict point or conflict segment, which will lead to another deadlock. Therefore, V x To obtain conflict resources, all AGVs with priority passage must have released the conflict resources.

[0056] S7 controls the waiting and waking of the AGV that actively avoids obstacles by setting a trigger.

[0057] The trigger is set up to ensure that the interlocked AGV can quickly wake up the AGV that is giving way after leaving the conflict section. The basic working principle of the trigger is to block the AGV when it reaches the avoidance point, and wake up the AGV that is giving way when the AGV with priority leaves the conflict section.

[0058] The triggers determine whether to allow the allocation of conflicting resources to V. x Have conflicting resources been released by the AGVs with priority passage?

[0059] In summary, in a one-way street scenario, when multiple AGVs are deadlocked, the optimal AGV is selected for active avoidance based on the set judgment conditions. Then, the route is planned using the selected avoidance points to find the AGV with priority passage. Finally, by setting triggers, the waiting and waking of the active avoidance AGV is controlled so that it can reach the target location and complete the task.

[0060] Example 2

[0061] Reference Figures 2-6 This is the second embodiment of the present invention, which differs from the first embodiment in that: two AGVs perform a route from the production area to the shelf area on a one-way road.

[0062] S1. Based on the directed graph formed by the current position and the next position of the AGV, allocate the execution route of the AGV.

[0063] like Figure 2 As shown in the diagram, the map in this example uses a one-way street as the main thoroughfare, connecting the production areas and shelving areas on both sides. To improve obstacle avoidance efficiency, several obstacle avoidance points, including A01, A02, and A03, are added to the one-way street based on the actual conditions of the workshop. Figure 2 PA and PB in the diagram represent the standby points of the AGV. Here, two AGVs are assigned to perform tasks and are named 0001 and 0002 respectively.

[0064] The route for car 0001 is PA-P01-P02-P04-P05-P07-P08-P09-P10-P11-P12-P14-P15-PL2; the route for car 0002 is PB-P15-P14-P12-P11-P10-P09-P08-P07-P05-P04-P03-P02-P01-PW2.

[0065] S2, determine if the directed graph is cyclic, check for deadlocks between AGVs, and denote the deadlocked AGV as V. i .

[0066] like Figure 3 As shown, traffic control occurs when car 0001 reaches P08 because P09 has already been assigned to car 0002. When car 0002 reaches P09, deadlock detection is performed. The two sets of information, P08 and P09 of car 0001 and P09 and P08 of car 0002, are used to generate a directed graph. The depth-first search method is used to check for cycles. At this point, it can be seen that a deadlock has been detected between car 0001 and car 0002.

[0067] S3. If a deadlock exists, the most suitable AGV is selected for active avoidance based on the comparison between the number of backoff steps and the number of nodes allowed to be avoided. The AGV that actively avoids the deadlock is denoted as V. x .

[0068] Taking two AGVs as an example, if the number of retreat steps for 0001 is less than that for 0002, 0001V1 is selected to avoid the obstacle; if the number of retreat steps for 0001 is greater than that for 0002, 0002 is selected to avoid the obstacle; if the number of retreat steps for both is the same, the AGV with fewer nodes allowed to avoid the obstacle is selected.

[0069] S4. According to the definition of allowed avoidance points, select the set of avoidance points P.

[0070] Perform a reverse search on the completed route portion of each interlocked AGV, checking the successor node of each point on the route.

[0071] The condition for a successor node to be a permissible avoidance point is as follows:

[0072] 1) The number of successor nodes is greater than 0.

[0073] 2) Not occupied by an idle AGV.

[0074] 3) It is not associated with any workstation, because workstations often contain important facilities such as equipment. In order to ensure compliance, this dispatch point should not be used as a pass point.

[0075] 4) Not on the subsequent route of the interlocked AGV.

[0076] like Figure 4 As shown, although P07 satisfies conditions 1 and 2, when car 0001 retreats to this point, it will still continue to lock up with car 0002.

[0077] Following the search method described above, car 0001 will search along P08-P07-P05-P04-P03-P02-P01-PW2, finding the first yielding point A03, requiring it to retreat 3 steps; car 0002 will search along P09-P10-P11-P12-P14-P15-PL2, finding the first yielding point A02, requiring it to retreat 4 steps. Therefore, car 0001 is chosen to yield.

[0078] S5, traverse the set of avoidance points P, use them as intermediate nodes, and plan two routes respectively.

[0079] In S3, the priority criterion for selecting an AGV is the number of retreat steps. During the selection phase of the optimal avoidance point, all avoidance points need to be re-compared, and the comparison criteria are as follows: total path cost and number of successor nodes. Avoidance points with lower total path cost are prioritized to improve overall efficiency; while avoidance points with 1 successor node have higher safety.

[0080] If there is no direct avoidance point, it is necessary to continue searching for an avoidance point along the subsequent route of the interlocked AGV; if no avoidance point is found, it is necessary to consider whether to move the AGV to the standby point for avoidance.

[0081] like Figure 5 As shown, car 0001 chose A03 as the optimal avoidance point. The complete route replanned by car 0001, including the completed route, is as follows:

[0082] PA-P01-P02-P03-P04-P05-P07-P08-P07-P05-A03-P05-P07-P08-P09-P10-P11-P12-P14-P15-PL2, where P07-P05-A03-P05-P07 is the turnaround route.

[0083] S6, the priority of AGVs that actively avoid obstacles is lower than that of AGVs that have priority to pass.

[0084] S7 controls the waiting and waking of the AGV that actively avoids obstacles by setting a trigger.

[0085] In a single task, there may be more than one conflicting path, and the waiting time for the avoidance AGV is not fixed in the actual scenario. In order to ensure that the interlocked AGV can be quickly woken up after leaving the conflicting path, a trigger needs to be set up.

[0086] The basic working principle of the trigger is to block the AGV when it reaches the avoidance point; and to wake up the AGV that actively avoids the conflict when the AGV with priority leaves the conflict section.

[0087] like Figure 5 As shown, after vehicle 0001 reached the yielding point A03, it entered a waiting state; as Figure 6 As shown, after car 0002 leaves P04, car 0001 is awakened at the avoidance point A03 and drives into P05. At this point, both cars are unlocked and continue to follow the replanned route to complete the mission.

[0088] In summary, in a one-way street scenario, when multiple AGVs are deadlocked, the best AGV is selected for active avoidance based on the set judgment conditions. Then, the route is planned using the selected avoidance points to find the AGV with priority passage. Finally, by setting triggers, the waiting and waking of the active avoidance AGV is controlled, thereby resolving the deadlock problem and reaching the target location.

[0089] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A single-lane scene-based multi-AGV unlocking method based on turn-back avoidance, characterized in that: The specific steps are as follows, A directed graph is formed according to the current position point and the next position point of the AGV, and the execution route of the AGV is distributed; Determine whether the directed graph is a loop, check whether there is a deadlock between AGVs, mark the AGVs in deadlock as V i ; If there is a deadlock, the number of steps to retreat and the number of nodes allowed to avoid are used as comparison conditions to select the most suitable AGV for active avoidance, and the AGV that actively avoids is denoted as ; According to the definition of the allowed avoidance point, the avoidance point set P is screened out; The avoidance point set P is traversed, and two route sections are planned respectively by taking the avoidance point as an intermediate node; The priority of the AGV that actively avoids is lower than that of the AGV that has priority to pass; The trigger is set to control the waiting and awakening of the AGV that actively avoids; The two segments include a route from the current location to an avoidance point and a route from the avoidance point to the destination node If the two segments exist, select the avoidance point with the smallest total route cost of and The priority of the AGV that actively avoids is less than the AGV that has the priority to pass includes when If the priority is higher when moving to the avoidance point, the original conflict point or conflict section resource is immediately obtained, which will cause another deadlock, therefore, The application to the conflict resource must satisfy the condition that the AGV that has the priority to pass releases all the conflict resources.

2. The method of claim 1, wherein the method further comprises: The number of AGVs includes at least two AGVs working simultaneously; the directed graph is connected to the production area and the shelf area according to the map model, taking a single lane as the main road; 3. The method of claim 2, wherein the method further comprises: determining whether the AGV is in a single-lane scene; and if the AGV is in a single-lane scene, determining whether the AGV is in a turning scene. The length of any path on the single lane can ensure that the two AGVs do not collide when they are located at both ends; on any path of the single lane, the AGV turning or rotating in place will not affect other AGVs.

4. The method of claim 3, wherein the method further comprises: The retreat step number refers to the number of steps required to search the avoidance point in reverse along the list of completed route sections in the current route; the allowed avoidance point refers to the scheduling node where the AGV is allowed to avoid; if the retreat step number is small, the AGV with a small retreat step number is selected to actively avoid; if the retreat step number is the same, the AGV with a small number of allowed avoidance nodes is selected.

5. The method of claim 4, wherein the method further comprises: The screening conditions of the avoidance point include that the number of successor nodes is greater than 0; it is not occupied by an idle AGV; it is not associated with any work station; and it is not on the subsequent route of the interlocked AGV.

6. The method of claim 5, wherein the method further comprises: From filtering out an allowed avoidance point from the direct successor node of the current node; If the set of avoidance points P is empty, then it is indicated that No direct avoidance is possible, the traversal V i of the unfinished route portion, if an avoidance point is found that allows avoidance, is added to P; If the avoidance point set P is still empty, the standby points defined in the map are traversed, and all nodes that have not been occupied by AGVs are added to the set P.

7. The method of claim 6, wherein the method further comprises: The trigger is set to ensure that the AGV that actively avoids can quickly wake up after the interlocked AGV leaves the conflict route section; the basic working mode of the trigger includes blocking the AGV that actively avoids when it reaches the avoidance point, and awakening the AGV that actively avoids when the AGV that has priority to pass leaves the conflict route section.

8. The method of claim 7, wherein the method further comprises: determining whether to allow the conflicting resource to be allocated to the AGV and the AGV with the priority of passage whether the conflicting resource is released.

Images