Output containment control of heterogeneous multi-agent systems with leaders of bounded inputs: An adaptive finite-time observer approach.