LSTM-Based Predictive Leader–Follower Control Scheme for Networked Drones
摘要
The multi-motor synchronization can be challenging due to variations in motor characteristics, propeller differences, and external disturbances. This paper suggests a two-tier control strategy for networked drones. At the local level, a quadruple Proportional–Integral–Derivative (PID) strategy has been proposed for such synchronization. Designing the optimal and predictive consensus control algorithm for networked drones is challenging. At a higher level, the paper proposes the long short-term memory (LSTM) based predictive and optimal leader–follower consensus control algorithm for the swarm of drones. The suggested predictive and converged consensus control scheme obtains the asymptotic steering of follower drones to leader drone.