Safety-Aware Cooperative Optimal Control for High-Altitude Cabin Air Extraction System: A Cooperative Differential Game and Control Barrier Function Approach

Conventional manual synchronization of the multiunit, multivalve air extraction system in high-altitude cabins suffers from operational inefficiency and poor disturbance rejection. The air extraction system is inherently safety-critical because compressor operation beyond design safety margins may induce catastrophic failure modes in high-altitude cabins. To guarantee both robustness and safety in the grid-connection process, this paper proposes a novel safety-aware cooperative optimal control architecture that integrates (1) cooperative differential game (CDG) theory for multiobjective optimization, (2) control barrier function (CBF) based cost function augmentation for formal safety guarantees, and (3) a composite controller combining active disturbance rejection control (ADRC) and proportional-integral (PI) control. This framework retains the performance benefits provided by CDG theory while ensuring formal safety guarantees via CBF throughout the learning phase. Rather than intervening only after constraint violations, the CBF formulation proactively co-optimizes safety and performance during planning, minimizing the need for corrective actions. Additionally, by integrating ADRC and PI algorithms in a composite architecture, it effectively resolves the dual challenges of low grid-connection efficiency and poor pressure control quality in the air extraction manifold. Extensive digital simulations under multiple operating conditions confirmed the reliability and effectiveness of the proposed approach. Compared with individual control, the proposed algorithm guarantees that the pressure ratio will remain within the safety margin throughout the entire control process, thereby ensuring system safety. This approach also results in a significantly shorter pressure setting time and a smoother dynamic response. The results demonstrate that the proposed framework outperforms traditional individual control in both effectiveness and safety, offering a robust solution for automated air extraction system.