Xiang Yu

In this paper, the water-entry behavior of an open cylindrical cavity is investigated by laboratory experiments and numerical simulations. Laboratory experiments are conducted in a water flume, with high-speed imaging used to capture cavity bubble evolution across a range of entry velocities, corresponding to Froude numbers Fr from 0 to 2.433. Three distinct flow patterns have been identified: a turbulent pattern (Fr1.75). The process is characterized by periodic liquid inflow and outflow, leading to repeated compression and expansion of the trapped gas, which in turn induces significant wall oscillations and morphological instability. A computational fluid dynamics model is utilized to further study the effect of the open-end bottom on the velocity and acceleration of the cavity. The results show that the open configuration experiences slower penetration, sustained fluctuations in velocity and acceleration, and enhanced unsteady dynamics due to internal fluid exchange.

For inertial measurement unit (IMU)/ ultra- wideband (UWB) integrated navigation systems, constructing the dynamics model based on additive errors is a common approach. However, the dependence of the error dynamics on the estimated navigation and bias states requires accurate initial state estimation, which may limit system application. To mitigate this limitation, several geometric navigation approaches have been proposed to decouple error dynamics from state estimation by introducing nonlinear error definitions. In this study, we firstly apply several nonlinear error representations, including two-frame group left- and right-invariant errors as well as tangent group errors to IMU/UWB integrated navigation to develop multiple distinct schemes. Furthermore, a novel left-invariant nonlinear error definition is proposed, wherein the error evolution of the navigation state achieves independent of navigation and bias estimation. The associated error dynamics and the integration scheme for the IMU/UWB-based navigation system are derived. The proposed methods are validated by using simulation data and a public dataset, demonstrating improved accuracy and reliability compared with the state-of-the-art approaches.

Accurate and robust state estimation, especially for attitude, velocity, and position, is crucial for robotics and autonomous navigation systems. This article introduces a novel robust navigation framework combining a geometric extended Kalman filter (GEKF) with dual robust kernels (DRKs). The GEKF reduces process model uncertainty by embedding navigation states into an extended special Euclidean group, addressing frame mismatch in error definition, and deriving a process model weakly dependent on attitude. The DRK technique enhances measurement update robustness by integrating convex and nonconvex kernels, with the convex kernel ensuring fast convergence and the nonconvex kernel suppressing outliers. Experimental results of the fusion between an Inertial Navigation System (INS) and Global Navigation Satellite System (GNSS) demonstrate that the proposed framework outperforms advanced methods in convergence speed, estimation accuracy, and robustness. This approach provides an innovative solution to improving navigation performance in complex environments.

For single-beacon navigation systems, the precision of the predetermined effective sound velocity (ESV) significantly affects the localization performance. However, the ESV between the pinger and receiver is time-varying, location-dependent, and difficult to determine accurately. ESV-setting errors induce large positioning errors and affect the practical applications of single-beacon navigation systems. Recently, unknown-ESV-based single-beacon navigation methods have been proposed to address the aforementioned issue, and their performances have been significantly better than those of known-ESV-based methods. This study aimed to conduct a theoretical analysis of the unknown-ESV-based single-beacon navigation methods in terms of observability. The local weak and global observabilities of an unknown-ESV-based single-beacon navigation system were investigated, and systems with and without ground-velocity measurements were considered. The observability of several common trajectories in underwater vehicles was summarized. Finally, a series of numerical simulations was conducted to validate the derived observability.

For an autonomous underwater vehicle (AUV) underwater localization system based on acoustic two-way travel time measurements, the motion of the AUV during the interrogation-reception time interval, unknown underwater sound velocity, and beacon calibration error affect the performance of the positioning solution. Ignoring these factors significantly deteriorates the performance of the localization system. This paper focuses on the localization problem of AUV in motion with an unknown sound velocity and beacon calibration error. The Cramér–Rao lower bound of the concerned localization problem is first derived as the performance bound of an unbiased estimator. Next, through multi-stage processing, we design a closed-form solution that can attain the corresponding performance bound under small measurement noise conditions. Furthermore, considering that the AUV motion velocity is typically much lower than the sound velocity in practical applications, a simpler and more efficient simplified closed-form solution is proposed, which can guarantee localization performance under the condition that the AUV motion speed is low, as well as small measurement noise. Finally, the theoretical performance and superiority of the proposed solutions were verified through numerical simulation.