Performance‐guaranteed containment control for pure‐feedback multi agent systems via reinforcement learning algorithm

Abstract

In this article, a performance-guaranteed containment control scheme based on reinforcement learning (RL) algorithm is proposed for a class of pure-feedback multi agent systems (MASs) with unmeasurable states. The unknown nonlinear functions are approximated by the neural networks (NNs) and an adaptive NN state observer is designed for the states estimation. Based on estimated states, the algebraic loop problem can be removed by introducing filtered signals, and the actor-critic architecture of RL algorithm is employed to acquire the optimal controller in the framework of backstepping. Different from many optimal strategies, this article proposes a simpler mechanism based on the uniqueness of the optimal solution to obtain the actor and critic updating laws instead of gradient descent algorithm with complicated calculation. In addition, predefined performance function and an improved error transformation technique are utilized to guarantee the containment error within a prescribed boundary. By using Lyapunov stability theory and graph theory, the stability of the closed-loop system can be demonstrated. Finally, the effectiveness of the method proposed in this article is verified by a simulation example.

Robust Distributed Containment Control with Adaptive Performance and Collision Avoidance for Multi-Agent Systems

“…The nonlinear filtering of the intermediate control signals prevents the complexity explosion and achieves convergence within a predefined time period. Alternatively, the work in [20] proposes a novel containment control scheme for multi-agent systems using reinforcement learning and neural networks. It addresses unmeasurable states with an adaptive observer and filtered signals.…”

Critic-Only Learning Based Tracking Control for Uncertain Nonlinear Systems with Prescribed Performance

“…For the high-order nonlinear multi-agent system containing uncertainty, the optimal consistency control problem with specified performance is considered in [26], where the stability of a closed-loop system and the convergence of consensus errors within a certain range are proved. Based on an actor-critic network, the optimal control is investigated for robot [27] and pure feedback system [28] separately. It is noted that the above controllers rely on multiple neural networks to ensure the stability and optimality of the system.…”

“…(2) Different from the optimal control derived by the actor-critic ADP framework in [24,[26][27][28], a critic-only NN was designed for online learn optimal control without constructing an actor NN. In addition, to achieve the convergence rate with a preassigned region, a novel value function was minimized, such that tracking errors can evolve within the prescribed region with low control consumption.…”