Achieving cooperative obstacle avoidance and formation control for multi-robot mobile systems in complex environments remains a key research challenge. Traditional centralized control approaches are vulnerable to single points of failure and computational bottlenecks, while distributed systems often suffer from poor formation stability. To address this, we propose a distributed control architecture that integrates a finite state machine, targeting cooperative obstacle avoidance and formation control in complex environments. The proposed framework drives state transitions based on a quantified conflict index and triggers leader switching when conflicts intensify. The new leader synchronizes the reference trajectory, and followers perform trajectory optimization based on their own perception. Simulation results demonstrate the effectiveness of the proposed method in ensuring stable obstacle avoidance and formation control for multi-robot systems in complex environments.