<p>This paper presents a control framework for a coaxial octocopter swarm operating in an environment with obstacle, addressing nonlinear dynamics and swarm coordination. First, the fully nonlinear dynamics of the coaxial octocopter are modeled. Then, LQR and model predictive controller are designed and applied to the nonlinear system, considering zero and nonzero yaw angle. However, when it is necessary to command the yaw direction specifically for big values, these controllers fail to manage the nonlinearities properly. To overcome this, we introduce a neuro-controller for online adaptation and precise yaw control, ensuring stability in the yaw angle. Due to its computational performance for real-word implementation, the LQR controller is extended into the Neuro-LQR to reach full stability in the system. Furthermore, we focus on multi-agent coordination for reconnaissance missions, where coaxial octocopters operate in a swarm and navigate obstacles. The coaxial octocopters synchronize their positions and yaw angles using consensus control systems, enabling coordinated obstacle avoidance and collective terrain scanning. This approach integrates robust control and synchronization, ensuring efficient swarm behavior. Sensitivity analysis of the system under varying initial conditions, conducted via Monte Carlo simulations, validates the robustness of the proposed framework. Results show that the control system ensures stable behavior of translational and rotational dynamics with minimal variability, demonstrating insensitivity to initial perturbations and highlighting the system’s ability to maintain reliable performance across missions.</p>