一种基于离散时间一致性理论的多机器人分布式巡逻算法

在多机器人巡逻任务中，由于通信距离的限制，单个机器人很难获得全局信息。然而，现有的大多数多机器人分布式巡逻算法都要求每个机器人获得其巡逻区域的全局信息进行决策。因此，考虑到通信半径约束和局部信息约束，为了通过相邻机器人之间的交互完成巡逻任务，基于离散时间一致性理论提出了两种巡逻算法。算法1使用全局信息进行决策，算法2基于离散时间一致性理论实现局部信息对全局信息的预测进行决策。通过模拟器Stage对所提算法与对比算法在不同机器人数量、通信半径、地图环境下进行了对比。实验验证了所提出的基于局部信息的分布式多机器人巡逻算法具有与原算法类似的特性和性能，能够使机器人在没有全局信息的情况下判断全局状态，并基于邻居之间的协商完成巡逻任务。

Distributed patrol algorithm for multi-robot systems based on discrete-time consensus theory

In the multi-robot patrol task, it is difficult for a single robot to obtain global information due to the limitation of the communication distance. However, most existing multi-robot distributed patrol algorithms require each robot to obtain global information of its patrol area for decision-making. Therefore, this paper addressed on the communication radius constrain and the local information constrain to completes patrol tasks through the interaction between neighbor robots, and proposed two solution algorithms based on discrete-time consensus theory. Algorithm 1 used global information for decision-making. Algorithm 2 used discrete-time consensus theory to predict and make decisions based on local information. Through the simulator Stage, the proposed algorithm and the comparison algorithm were simulated and compared under different robot numbers, communication radii and map environments. The simulation verifies that the proposed fully distributed multi-robot patrol algorithm based on local information has similar characteristics and performance to the original algorithm, enabling robots to judge the global state without global information and complete patrol tasks based on the negotiation between neighbors.