关键词:
Adaptive dynamic programming
Output constraints
Transformation function
Neural network observer
摘要:
A state consensus cooperative adaptive dynamic programming (ADP) control strategy is proposed for a nonlinear multi-agent system (MAS) with output constraints. On the basis of the transformation function, state models of leader and followers are transformed into affine ones. By using a monotonically increas-ing mapping function, the state-consensus cooperative control problem for an MAS with output con-straints is equivalently transformed into a cooperative approximately optimal control one for an affine MAS. Then, a neural network observer is constructed for estimation of inner states, and, by graph theory and ADP method, the state consensus cooperative ADP control strategy is developed. The proposed strat-egy guarantees the performance index of the transformed system is approximately optimal. Furthermore, the stability analysis of whole closed-loop system is presented. Through the Lyapunov Theorem, we prove that the states of the MAS achieve consensus and the output signals of the followers satisfy the con-straints. Also, all signals of the closed-loop MAS are bounded, and the trajectory of the leader node is cooperative bounded. The theoretical analysis and effectiveness of the strategy are verified by both a physical and a numerical example. (c) 2021 Elsevier B.V. All rights reserved.