A free-swimming tadpole model based on immersed boundary-lattice Boltzmann method and its application

The autonomous swimming of fish in a complex flow environment is a nonlinear and intricate system, which is the focus and challenge in various fields. This study proposed a novel simulation framework for artificial intelligence fish. It employed a high-precision immersed boundary-lattice Boltzmann coupling scheme to simulate the interactions between fish and flow in real time, and utilized the soft actor-critic (SAC) deep reinforcement learning algorithm for fish brain decision-making module, which was further divided into a vision-based directional navigation and a lateral line-based flow perception modules, each matched with its corresponding macro-action space. The flow features were extracted using a deep neural network based on a multi-classification algorithm from the data perceived by the lateral line and were linked to the fish actions. The predation swimming and the various Kármán gait swimming were explored in terms of training, simulation, and generalization. Numerical results demonstrated significant advantages in the convergence speed and training efficiency of the SAC algorithm. Owing to the closed-loop “perceive-feedback-memory” mode, intelligent fish can respond in real-time to changes in flow fields based on reward-driven requirements and experience, and the accumulated experience can be directly utilized in other flow fields, and its adaptability, model training efficiency, and generalization were substantially improved.

A numerical simulation research on fish adaption behavior based on deep reinforcement learning and fluid–structure coupling: Implementation of the “perceive-feedback-memory” control system

Zhang,

Li,

Zhang

et al. 2024

Surface hydrodynamic characteristic analysis of underwater robots applied to perception of lateral lines

Li,

Guo

2024

Inspired by the hydrodynamic perception abilities based on lateral lines on fish surfaces, the artificial lateral lines (ALLs) based on pressure and flow sensors were proposed by the researchers. As the ALLs are widely used in underwater robots, the mechanisms of lateral line perception are urgently needed to be studied. Based on the lattice Boltzmann method, immersion boundary method, and large eddy simulation, a three-dimensional numerical model of underwater robot motion is established and verified. The distribution and variation of velocity and surface pressure on robots with different shapes under different flow fields are studied in detail. It is found that the robots with the upstream surface curvature aspect ratio of 1:1 are more suitable for placing ALLs. Then, similarly, the hydrodynamic perception abilities of robots with different sizes are further investigated. It was observed that the smaller the robot size, the better the perception ability. In addition, sensing devices are more suitable for placement on the upstream surface of the robots. These conclusions can also explain the physiological characteristics of cavefish with well-developed lateral lines in nature. Finally, based on the above analysis, to guide the shape design and sensor layout of the robots, an evaluation index for the perception ability of the robot is proposed. The reliability of the evaluation index is verified by using a machine learning method based on polynomial regression to predict the flow field. The R-square of machine learning can reach 0.99 at the upstream surface of the robot.

Energy conservation analysis and formation control of multi-tadpoles based on lattice Boltzmann method

Li,

Xu,

2024

Tadpole swimming is a common swimming form in nature, but there are only few studies on the swimming properties of tadpoles and tadpole groups. This paper focuses on the control and energy consumption during tadpole swimming and studies the swimming formation suitable for the tadpole group based on their hydrodynamic properties. First, based on the immersed boundary-lattice Boltzmann method and tadpole model proposed in our previous studies, the swimming control strategy of tadpoles based on incremental proportional–integral–derivative control is proposed. Then, the influence of control quantity on tadpole swimming is analyzed, and the control thresholds are set to realize the stable swimming control of tadpoles. Furthermore, through energy analysis of the pairs of tadpoles at different lateral distances, phase differences, and horizontal distances, we found the reduction of tadpole lateral distance has a slight effect on energy-savings. It is the energy-efficient mode for swimming pairs of adjacent tadpoles, when the longitudinal distance is one body length and the phase difference is 5π/3. In addition, under different longitudinal distances, the phase difference for energy efficiency of tadpoles is also different, but it shows a periodicity. Finally, the study results were extended to a formation of three or four tadpoles for verification. Our research will provide a mechanical reference for studying tadpole swarm swimming and for the design and control of tadpole-like robots.