Research On Environmental Perception And Formation Behaviors Of Bionic Fish Swarm Based On Numerical Simulation

With the development of ocean exploration,bionic fish has become a hot topic in the field of underwater robots because of its superior performance.The bionic fish swarm can be more flexible,more efficient and do more complex tasks than single bionic fish.However,with a bionic fish swarm,unlike air or land robots,it is difficult to obtain environmental information through traditional means such as communication and acoustic systems.And now there is no unified conclusion about the efficient formation of the bionic fish swarm.Aiming at the above problems,we use the simulation method to study the environmental perception and formation behaviors of the bionic fish swarm.The main work and innovations are as follows:(1)We build a simulation platform of bionic fish based on the open-source CFD software Open FOAM.According to the three processes of CFD,we introduce the preprocessing of geometric model,kinematic model and mesh generation method of bionic fish,introduce the key technologies in the solution process: an arbitrary Lagrangian-Eulerian(ALE)method and a parallel dynamic mesh method based on the radial basis function,introduce the post-processing ideas of simulation results,and present a platform foundation for the follow-up research of biomimetic fish.(2)We build the environmental perception model of the bionic fish swarm.In this paper,starting from the needs and research methods of environmental perception,we design the framework of environmental perception.We design the sensor data extraction algorithm corresponding to the bionic fish's geometric model and kinematic mode.We analyze the sensitivity of different sensors distributed on the bionic fish's surface to the velocity and spacing and find that sensors located from the head to about one-third point of the front body are more useful.The result can provide guidance for sensor deployment in physical environment experiment.For different bionic fish formation,we select the sensor data in the appropriate position,use the machine learning method to establish the perception model,and achieve high accuracy.(3)We find the efficient formation of the bionic fish swarm.In this paper,we design four formations(the tandem,the phalanx,the diamond,and the rectangle)of bionic fish,and the performance of bionic fish formations are measured by Froude efficiency that indicates the proportion of the useful power in the total power.Based on the analysis of the velocity and pressure fields,and comparing the Froude efficiency of different formations,we find the formation of high-efficiency: when the fish spacing is small,the tandem swarm is the best choice.When the spacing is large,the rectangle swarm is more efficient than other formations.