To address the collaborative obstacle avoidance problem for multi-aircraft in complex dynamic environments under model uncertainty and unknown disturbances, this paper proposes a dynamic adaptive cooperative obstacle avoidance method that integrates the Centroidal Voronoi Tessellation (CVT) algorithm with a fixed-time consensus strategy. This approach aims to overcome the shortcomings of traditional obstacle avoidance algorithms, such as insufficient real-time responsiveness and poor environmental adaptability in multi-aircraft systems. By combining the spatial partitioning characteristics of the CVT algorithm with the distributed rapid coordination mechanism of the fixed-time consensus strategy, an efficient dynamic obstacle avoidance framework is established. First, the CVT algorithm is employed to partition the task space into multiple dynamic sub-regions, and the boundaries of these regions are optimized in real time based on aircraft motion states to achieve global obstacle avoidance path planning. Second, an enhanced consensus protocol is designed to coordinate the motion states among multi- aircraft, ensuring fixed-time consistency between individual objectives and group obstacle avoidance constraints, thereby effectively mitigating local oscillations and collision risks. Additionally, to address potential collisions caused by unknown disturbances and model parameter uncertainties during multi-aircraft obstacle avoidance, fuzzy logic systems (FLSLs) are introduced to approximate uncertain parameters and unknown interference terms in real time, further enhancing aircraft safety. Simulation results demonstrate that the proposed method not only guarantees multi-aircraft safety but also exhibits superior adaptability to dynamic environments and faster convergence rates, highlighting its significant engineering application value.