Cooperative Control of Multiple Agents with Unknown High-frequency Gain Signs under Unbalanced and Switching Topologies
For researchers in multi-agent control, this work relaxes restrictive topology assumptions (balanced, strongly connected) to more realistic unbalanced and switching conditions, though the improvement is incremental.
This paper proposes nonlinear PI-based algorithms for cooperative control of multi-agent systems with unknown high-frequency gain signs, enabling operation under unbalanced and switching topologies with a jointly strongly connected basis, whereas prior work required fixed or switching strongly connected and balanced graphs. The approach is validated for single-integrator and double-integrator agents via simulations.
Existing results on cooperative control of multi-agent systems with unknown control directions require that the underlying topology is either fixed with a strongly connected graph or switching between different strongly connected graphs. Furthermore, in most cases the graph is assumed to be balanced. This paper proposes a new class of nonlinear PI based algorithms to relax these requirements and allow for unbalanced and switching topologies having a jointly strongly connected basis. This is made possible for single-integrator (SI) and double-integrator (DI) agents with non-identical unknown control directions by a suitable selection of the distributed nonlinear PI functions. Moreover, as a special case, the proposed algorithms are applied to strongly connected and fixed graphs. Finally, simulation examples are given to show the validity of our theoretical results.