Practical bipartite consensus for multi-agent systems: A barrier function-based adaptive sliding-mode control approach