Swarm escape and convergence control based on distributed intent recognition under network attack
When a swarm composed of multiple robots or drones detects an unknown external node during the execution of a task,it needs to identify its intention to decide how to respond.However,the internal nodes of the swarm will be attacked against the measurement signal,resulting in increased measurement error,and then the intention identification of the external nodes will fail,posing a threat to the swarm.To solve this problem,a distributed control algorithm for swarm which considers the network attack is designed.In this algorithm,the control law is the flocking algorithm.When an unknown external node is detected,firstly,the internal node determines whether it is under network attack based on the designed attack recognition algorithm.Then,according to the distributed Kalman filtering based on attack identification strategy,the state of external nodes is estimated to minimize the impact of network attacks.Next,the similarity between the expected trajectory and the measured trajectory is calculated according to the Fréchet distance,and the distributed consensus algorithm is used to judge the intention of the external node,control swarm escape or converge.Finally,the effectiveness of the proposed method is demonstrated by simulation results.
NETL
NSTL
万方数据
