CN-121995873-A - Dynamic master-slave system and election method for multi-AGV cooperative scheduling

CN121995873ACN 121995873 ACN121995873 ACN 121995873ACN-121995873-A

Abstract

The invention relates to an AGV scheduling system and method, in particular to a multi-AGV cooperative scheduling dynamic master-slave system and an election method, which are used for solving the problems that a dynamic master node election mechanism is lacked in the existing distributed AGV group, the single-point failure recovery of a master node is slow, and the roles of AGVs cannot be dynamically switched. The invention comprises a global state synchronization module, a scheduling decision module, a data storage module, a dynamic role management module, a wireless communication module and a heartbeat monitoring and fault processing module which are arranged on each AGV. And driving each AGV to participate in election based on a multivariate evaluation function through triggering an event, electing a master node through distributed negotiation consensus, and registering the rest as slave nodes. The master node periodically broadcasts a heartbeat signal, the slave nodes cooperatively monitor, and the reelect is triggered once the failure of the master node is confirmed. The invention realizes the autonomous organization of the decentralized AGVs, the fault self-recovery and the role dynamic switching, and remarkably improves the reliability, the expansibility and the self-adaptive capacity of the multi-AGV system.

Inventors

LIU BO
HAN HAIWEI
ShangGuan Qingyun
XU HAOCHENG
Jiang Shunchang
ZHAO MING

Assignees

西安航天赛能自动化科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251225

Claims (10)

1. The multi-AGV collaborative scheduling dynamic master-slave system is used for dynamically selecting a master node AGV and a slave node AGV for an AGV vehicle group formed by a plurality of AGVs, and is characterized by comprising a global state synchronization module, a scheduling decision module, a data storage module, a dynamic role management module, a wireless communication module and a heartbeat monitoring and fault processing module which are arranged on each AGV; In each AGV, the dynamic role management module is connected with the wireless communication module, and the heartbeat monitoring and fault processing module is respectively connected with the wireless communication module and the dynamic role management module; the system comprises a data storage module, a global state synchronization module, a scheduling decision module, a dynamic role management module, a task execution entity and a task execution entity, wherein the data storage module is used for storing data of the AGVs, the global state synchronization module is used for storing the data of the AGVs, and the global state synchronization module is connected with the wireless communication module; The system comprises a master node AGV, a global state synchronization module, a scheduling decision module and a scheduling decision module, wherein the master node AGV activates the global state synchronization module and the scheduling decision module; the scheduling decision module is used for generating scheduling instructions according to the global state view and scheduling rules prestored in the data storage module, and issuing the scheduling instructions to corresponding slave nodes AGVs through the wireless communication module; the heartbeat monitoring and fault processing module in the master node AGV is used for periodically broadcasting heartbeat signals to the slave node AGVs, each slave node AGV center hop monitoring and fault processing module is used for monitoring the periodic heartbeat signals broadcast by the master node AGVs so as to judge whether the master node AGVs are faulty or not, and a master-slave election instruction is triggered when the master node AGVs are monitored to be faulty.
2. The multi-AGV collaborative scheduling dynamic master-slave system according to claim 1 wherein each dynamic role management module comprises a parameter calculation unit, a message receiving and transmitting unit and a decision logic unit; each parameter calculation unit is used for calculating the election parameters of the AGV based on a preset multi-element evaluation function when election is triggered; Each message transceiver unit is connected with a corresponding wireless communication module and is used for broadcasting an election message containing the election parameters through a wireless network and receiving the election messages from other AGVs; Each decision logic unit is connected with the corresponding parameter calculation unit and message receiving and transmitting unit and is used for comparing the election parameters in all the received election messages and determining a new main node AGV according to a preset winning or losing judgment rule; Each heartbeat monitoring and fault processing module comprises a timeout counter, a state cooperative verifier and a fault determiner; Each timeout counter is used for counting when no valid heartbeat signal is received in a preset heartbeat period; each state cooperative verifier is connected with a corresponding overtime counter and a wireless communication module and is used for exchanging state confirmation information with other slave nodes AGVs through a wireless network when the count value of the corresponding overtime counter reaches a threshold value; And each fault determiner is connected with the state collaborative verifier and is used for generating the fault trigger signal after receiving more than half of other slave node AGVs confirm the suspicious information of the master node AGVs.
3. A dynamic master-slave election method for multi-AGV cooperative scheduling is characterized by comprising the following steps: s1, constructing a multi-AGV cooperative scheduling dynamic master-slave system according to claim 1 or 2; S2, presetting a task target to be completed, powering up all AGVs in the AGV vehicle group, and triggering a master-slave election process of the AGV vehicle group after the AGVs are initialized; S3, each candidate AGV calculates self-election parameters according to a preset multivariate evaluation function through a dynamic role management module, and broadcasts an election message containing the election parameters into an AGV vehicle group; s4, each AGV receives and compares the competition information broadcast by all AGVs through the dynamic role management module, and each AGV autonomously achieves consensus about the identity of the AGV of the unique master node according to a preset winning or losing judgment rule; S5, the AGVs which are determined to be the master nodes by the consensus activate a global state synchronization module and a scheduling decision module, and broadcast the identities of the master node AGVs through a dynamic role management module; s6, a scheduling decision module of the master node AGV generates a scheduling instruction according to the global state view generated by the global state synchronization module and a scheduling rule prestored in the data storage module, the scheduling instruction is issued to a corresponding slave node AGV through the wireless communication module, the slave node AGV is scheduled to execute a related operation task, and heartbeat signals are periodically broadcast to the slave node AGV through the heartbeat monitoring and fault processing module; And S7, the slave node AGVs monitor the periodical heartbeat signals broadcast by the master node AGVs through the heartbeat monitoring and fault processing module, judge whether the master node AGVs are faulty or not, and if the master node AGVs are faulty, return to the step S3 to finish the election of the new master node AGVs again until a preset task target is completed, so that the dynamic master-slave election of the multi-AGVs cooperative scheduling is realized.
4. The method for dynamically selecting master and slave for coordinated scheduling of multiple AGVs according to claim 3, wherein step S7 specifically comprises: S7.1, each slave node AGV monitors a periodical heartbeat signal broadcast by the master node AGV, judges whether the heartbeat signal of the master node AGV is received or not through a heartbeat timeout timer arranged in the slave node AGV, and marks the master node AGV as suspicious locally in the AGV if the heartbeat message of the master node AGV is not received in N continuous heartbeat periods, wherein N is more than or equal to 2; s7.2, any slave node AGV marking the master node AGV as a suspicious state broadcasts a query request message containing the suspicious state of the master node AGV into an AGV vehicle group; s7.3, each slave node AGV responds to the query request message, and if the slave node AGVs also mark the master node AGVs as suspicious, the slave node AGVs reply to the confirmation message; And S7.4, if the slave node AGVs initiating the query receive more than half of the confirmation information replied by other slave node AGVs within the preset time, cooperatively judging the fault of the master node AGVs, returning to the step S3 to finish the election of the new master node AGVs again until the preset task target is completed, and realizing the dynamic master-slave election of the multi-AGV cooperative scheduling.
5. The method for dynamically selecting master and slave for coordinated scheduling of multiple AGVs according to claim 3, wherein step S4 further comprises an authentication and anti-collision mechanism for the election message, specifically: the election message comprises a digital signature or a message verification code of a sending node; After receiving the message, the receiving node firstly verifies the validity of the signature and only processes the message passing verification; if the same AGV receives a collision announcement from a different node, but claims to be the same node, then only the latest announcement is accepted according to the timestamp or sequence number in the message.
6. The method for dynamically selecting master and slave for coordinated scheduling of multiple AGVs according to claim 5, wherein in step S4, the predetermined win/lose determination rule is specifically: Each AGV compares the competitive parameters in all the received competitive messages, and determines the AGV with the highest competitive parameter as a main node AGV; if the competitive parameters are the same, determining according to a preset priority strategy, wherein the priority strategy comprises comparing the size of a physical identification ID of the AGV or comparing the latest update time stamp of the AGV.
7. The method for dynamic master-slave election with multi-AGV cooperative scheduling according to claim 3, wherein in step S5, the registration information sent when the slave node AGV registers with the master node AGV at least includes: The unique identifier of the node, the current task state, the current fault state, the current location information, the current power state, the current load capacity, and the list of executable task types.
8. The method for dynamic master-slave election with multi-AGV cooperative scheduling according to claim 3, wherein in step S3, the preset multiple evaluation function is a weighted function, and the variables include at least: The percentage of the remaining power of the AGV, the current load rate of a central processing unit of the AGV, the average communication signal intensity with other AGVs in the vehicle group and the historical continuous normal working time.
9. The multi-AGV co-scheduled dynamic master-slave election method of claim 8, wherein: In step S3, the weight parameter of the multivariate evaluation function performs offline or online learning adjustment according to the historical operation data of the vehicle group, so as to optimize the influence of the election result on the overall operation efficiency of the system.
10. The dynamic master-slave election method of multi-AGV coordinated scheduling according to any one of claims 3-8, wherein: In the step S4, a first overtime is also set in the consensus reaching stage, if the first overtime expires, the AGVs still do not reach consensus for the identity of the AGVs of the master node, the current election state is cleared, and the step S3 is returned to initiate election again; In step S5, a second timeout period is set in the identity confirmation stage of the master node AGV and the slave node AGVs, and if the master node AGV does not receive the registration information of all the slave node AGVs before the expiration of the second timeout period, the master node AGV broadcasts a registered slave node AGV list to an AGV vehicle group, the unregistered slave node AGVs are regarded as offline, and then the slave node AGVs can be added into the system through a re-registration process.

Description

Dynamic master-slave system and election method for multi-AGV cooperative scheduling Technical Field The invention relates to an AGV scheduling system and a method, in particular to a dynamic master-slave system and a dynamic master-slave election method for multi-AGV cooperative scheduling. Background Along with the wide application of AGVs in flexible production lines, intelligent storage and other scenes, a multi-AGV collaborative operation system is evolving from a traditional centralized architecture to a distributed and decentralized architecture. In the centralized architecture, the system relies on a single control server for task scheduling, and the problems of high single-point fault risk, high network deployment and maintenance cost, limited system expansibility and the like exist. In the prior art, a scheme for realizing hard synchronization of multiple AGVs through physical connection is available, for example, china patent CN119937539B discloses a system and a method for controlling multiple AGVs in a coordinated manner. The system is composed of a plurality of independent AGV electrical control systems, each system comprises a controller, a remote controller, a driver and a motor, and the controller is connected with the front AGV system, the rear AGV system, the driver and the remote controller through various interfaces to realize movement level linkage. However, such schemes are essentially extensions of motion control and cannot support complex logical task collaboration and dynamic system management. Other researches in the prior art propose a multi-AGV scheduling method, but the core problem of how to realize a highly reliable and self-organizing node election and scheduling mechanism in a distributed AGV vehicle group without a fixed center server is not deeply solved. In particular, in a fully distributed AGV system, the critical ability to dynamically elect the master node that assumes the centralized scheduling responsibilities remains lacking. And secondly, how to realize rapid and collaborative fault detection and system self-recovery when the main node fails. Thirdly, how to enable the common AGV to have role dynamic switching capability and support the system to reconstruct a logic architecture in operation. Disclosure of Invention The invention aims to solve the technical problems that a dynamic master node election mechanism is lacked in the existing distributed AGV group and the AGV roles cannot be dynamically switched, and provides a multi-AGV cooperative scheduling dynamic master-slave system and an election method. In order to achieve the above purpose, the invention adopts the following technical scheme: The multi-AGV collaborative scheduling dynamic master-slave system is used for dynamically selecting a master node AGV and a slave node AGV for an AGV vehicle group formed by a plurality of AGVs, and is characterized by comprising a global state synchronization module, a scheduling decision module, a data storage module, a dynamic role management module, a wireless communication module and a heartbeat monitoring and fault processing module which are arranged on each AGV; In each AGV, the dynamic role management module is connected with the wireless communication module, and the heartbeat monitoring and fault processing module is respectively connected with the wireless communication module and the dynamic role management module; the system comprises a data storage module, a global state synchronization module, a scheduling decision module, a dynamic role management module, a task execution entity and a task execution entity, wherein the data storage module is used for storing data of the AGVs, the global state synchronization module is used for storing the data of the AGVs, and the global state synchronization module is connected with the wireless communication module; The system comprises a master node AGV, a global state synchronization module, a scheduling decision module and a scheduling decision module, wherein the master node AGV activates the global state synchronization module and the scheduling decision module; the scheduling decision module is used for generating scheduling instructions according to the global state view and scheduling rules prestored in the data storage module, and issuing the scheduling instructions to corresponding slave nodes AGVs through the wireless communication module; the heartbeat monitoring and fault processing module in the master node AGV is used for periodically broadcasting heartbeat signals to the slave node AGVs, each slave node AGV center hop monitoring and fault processing module is used for monitoring the periodic heartbeat signals broadcast by the master node AGVs so as to judge whether the master node AGVs are faulty or not, and a master-slave election instruction is triggered when the master node AGVs are monitored to be faulty. Further, each dynamic role management module comprises a parameter calculation unit, a message