The complexity and variability of the deep-sea environment present significant challenges for autonomous underwater robots, particularly in dynamic modeling considering environmental disturbances. This paper presents a novel environment uncertainty-aware dynamic modeling approach for autonomous deep-sea robots. First, a robot multibody dynamic model is established under ideal environmental conditions. Then, environmental disturbances, including pressure, temperature, and density, are incorporated to capture their environment–robot coupling effects. Finally, a neural network compensator is designed to predict pose deviations caused by uncertain ocean disturbances. Experimental studies under deep-sea conditions show that the proposed model can accurately predict the motion state of the deep-sea robot with most depth errors within 5 m and pitch angle errors within 0.1 rad, providing a solid foundation for future autonomous deep-sea robot motion control and autonomous navigation.

In this paper, we present in this work a fairly complete process for developing an unmanned aerial–aquatic vehicle system, TJ-FlyingFish, which includes an innovative design methodology of the aerial–aquatic platform and the cross-medium localization, dynamics modeling, and flight control systems. The development faces the challenge how to manipulate locomotion effectively in both water and air which presents substantial differences in fluid properties. Additionally, there are difficulties in perception and navigation because of the discontinuity of mediums. To cope with these challenges, we designed an innovative unmanned aerial–aquatic vehicle with an optimized dual-speed and tilting propulsion configuration. The rotors/propellers operate in different ranges of rotating speed in the two different mediums, providing sufficient thrust and ensuring output efficiency. Besides, thrust vectoring is achieved by rotating each propulsion unit around its mounted arm, facilitating agile underwater cruising. Another key component of our approach is a sophisticated multi-sensor-based cross-medium localization system that combines SLAM, sensor synchronization, and data capture mechanisms, enabling seamless transitions between aerial and aquatic environments, and supporting autonomous operations. The results are fully validated through actual flight experiments.