Cooperative Adaptive Formation Fault-Tolerant Neural Control for Multiple Quadrotors With Full-State Constraints

Abstract

This paper investigates the cooperative time-varying formation fault-tolerant control problem for multiple quadrotors with unknown actuator faults and full state constraints. In order to ensure the safety and operability of quadrotors in the confined flight environment, a novel transformed function is first introduced to convert the original quadrotor systems into unconstrained equivalent systems, which increases the flexibility of the controller design. Then, a distributed kinematic control protocol and fault-tolerant dynamic control protocol using the adaptive neural networks estimation technique are developed to guarantee the cooperative time-varying formation of multiple quadrotors subject to uncertain parameters. Meanwhile, the unknown actuator loss of effectiveness and bias faults are compensated and the state variables of position subsystem and attitude subsystem can be maintained within the designed performance constraint sets even when actuator faults occur. Via Lyapunov stability theory, the cooperative formation fault-tolerant performance analysis is presented. The proposed control strategy is validated through simulations.