Cavitation Dynamics and Control Strategy for High-Speed Water Entry of Trans-Media Vehicle

Accurate trajectory control during the high-speed water entry of trans-media vehicles faces significant challenges due to complex cavitation dynamics and highly nonlinear fluid-structure interactions. This paper proposes a robust underwater trajectory stabilization strategy leveraging a Linear Quadratic Regulator (LQR) controller. First, a computational hydrodynamic model is established using the slicing method: the axisymmetric vehicle geometry is discretized into longitudinal segments, and time-varying wetted regions are dynamically determined by analyzing the geometric interaction between the expanding/contracting cavity and the vehicle surface. This model, integrated with six-degree-of-freedom rigid-body motion equations, enables efficient full-trajectory dynamic simulation from cavity navigation to fully wetted phases. Second, an optimal LQR controller is designed to generate horizontal, vertical, and differential rudder deflection commands. A saturation function explicitly constrains rudder angles within physical actuator limits, ensuring feasible control inputs while achieving stable underwater attitude regulation. Finally, comprehensive simulations validate the controller's performance across diverse entry conditions. Key results demonstrate: Compared to a conventional PID controller, the LQR achieves an 81.29% improvement in the ITAE (Integral of Time multiplied by Absolute Error) metric for depth tracking under high-speed/large-angle entry (150 m/s, 90°), reducing pitch angle convergence time to under 2 seconds. At moderate entry conditions (100 m/s, 60°), depth control ITAE improves by 60.95%. The LQR also exhibits significantly faster rudder response, effectively suppressing depth overshoot (15.31mvs. PID's 4.71 m in high-speed case). This strategy provides a reliable and efficient solution for depth-keeping and attitude stabilization of trans-media vehicles operating in rapidly evolving hydrodynamic environments.