Distributed Fixed-Time Prescribed Performance Consensus of Heterogeneous Nonlinear Multi-Agent Systems with Mismatched Disturbances

Abstract

This paper studies distributed leader–follower consensus for heterogeneous nonlinear multi-agent systems with different physical orders and channel-dependent matched and mismatched disturbances. A channel-wise fixed-time extended-state observer estimates the lumped disturbance in each state equation, while a finite-terminal-time performance function constrains neighborhood errors during the transient and steady-state stages. Order-dependent composite surfaces and controllers are designed without augmenting lower-order followers to a common order. Sufficient conditions are derived for fixed-time observer convergence, practical fixed-time convergence of the composite surfaces, boundedness of the complete closed loop, and invariance of the prescribed-performance funnels. Simulations with followers of orders [2,3,2,3] demonstrate disturbance reconstruction, prescribed-performance satisfaction, bounded control effort, and reduced sensitivity of the dominant observer–surface convergence phase to the initial condition.