Long- and short-term dynamic risk cooperative control of deepwater drilling riser driven by an uncertainty-based mechanical model

The deepwater drilling riser is structurally fragile, facing both long- and short-term risks that are difficult to manage simultaneously in complex marine environments. Existing strategies focus on single-event control via displacement adjustment, lacking a coordinated framework for multi-timescale risk mitigation. To fill this gap, we propose a cooperative control method driven by an uncertainty-based mechanical model. A risk assessment model is established first. For long-term control, a feedforward approach classifies potential faults and builds a preventive rule base with predefined trigger times, whose execution resets fault occurrence rates. For short-term control, a feedback method uses top and bottom flexible joint angles as optimization objectives; ant colony optimization tunes tension and platform position under various sea states, and a Bayesian-optimized support vector regression constructs a data-driven controller, with an intermittent control criterion to save costs. The two strategies are integrated to handle long- and short-term risks concurrently. Results demonstrate that our method substantially reduces long-term risk and expands the short-term safe sea state envelope. This work provides a practical solution by synergizing knowledge-based feedforward rules with state-regulated feedback control for enhanced riser safety under uncertainty.