Iterative Learning Control of Multi-Agent Systems under Changing Network Configuration

This paper develops a new distributed iterative learning control (ILC) design method for uncertain multi-agent systems under changing network configuration between passes. It is assumed that the network configuration is not fixed, but the set of possible configurations is known in advance; the network configurations can differ in both the number of agents and the connections between them, and the transition from one configuration to another occurs according to an external logical rule. In particular, these assumptions reflect smart manufacturing features when the technological process must be quickly rebuilt. For reducing the transient error when changing the network configuration, a special ILC switching rule is proposed. Copyright (C) 2021 The Authors.