Flow-sensing-in-loop Adaptation Control For Bio-inspired Underwater Robots

Fish are able to adapt for the process of natural selection by developing a distinctive and remarkable ability of movement in the water. Lateral line plays an important role as the key component of fish perception system. Fish can sense the slow evolution of flow field surrounded by sensing organ, so as to keep sustained speed under low energy consumption and high maneuverability. Inspired from fish sensing their external flow, flow information featured by near-body pressure is of significance to enable underwater locomotion controllers with more adaptability and efficiency within varying environments.By fusing the theoretical, computational and experimental methods, the dissertation concentrates on flow sensing based control of bio-inspired underwater robots, with the following achievements and progress.(1) The complicated coupling process between the fluid environment and the dynamic behavior of fish is analyzed. Computational fluid dynamics(CFD) is employed into unveiling the bio-inspired fish swimming hydrodynamics. This dissertation proposes and develops a simulation environment for understanding fish locomotion and hydrodynamic effects during the self-propulsion in flow fields. Approximate kinematics model and profile description based on camera observation are recommended to specify active fish deformation. The CFD calculation using base-fixed or behavior-coupled model is executed. This work has formed a meaningful basis of computational platform for further researches on bio-inspired propulsion, by constructing virtual flow information perception, locomotion feedback and controller module.(2) The external flow field and its own motion effect on the near-body pressure are analyzed. The flow sensing and estimation method is proposed and implemented. The relationships between surface pressure coefficient and flow filed parameters are established, in both steady uniform flow and unsteady Karman vortex street based on theoretical analysis and CFD numerical simulation. The effect of self-motion on nearbody pressure is fully discussed during self-propelled swimming. Furthermore, a filtering algorithm is designed and implemented to fuse near-body pressure of one or multiple points for the estimation of the external flow. The simulation results demonstrate that the proposed computational scheme and its corresponding algorithm are both effective to predict the inlet flow velocity with near-body pressure at specified spatial points.(3) The cruising efficiency evaluation is established, and the propulsion performances are comparatively analyzed of the Anguilliform and Carangiform modes. The results show that there are multiple options of wave motion parameters to achieve desired cruising velocity, while the efficiencies appear different. Wave motion parameter of optimized efficiency is preferred as an adaptive reference value for online motion control. The cruising velocity increases monotonically with the wave frequency and amplitude. The cruising efficiency decreases as the wave frequency increases, and it firstly increases and then decreases with the increasing wave amplitude. The bio-inspired fish swimming with flexible tail has better adaptability than rigid tail, while the cruising speed is roughly equivalent, but the cruising efficiency is improved by 58%.(4) Flow sensing based adaptation control method is proposed to improve the control accuracy and swimming performance. Flow sensing and adaptive reference value are introduced into the control loop to solve the problem of station holding. Flow sensing is beneficial for control convergence by decreasing the convergence time. The adaptive iterative learning control method is used to further improve the accuracy, since it is adaptive for time-varying parameters and modeling uncertainty.(5) A flow sensing robotic fish prototype is designed and implemented to verify pressure sensing based adaptation control method. Near-body pressure is sensed and further used for estimation. Self-propelled cruising locomotion and efficiency evaluation are fully analyzed. When the robotic fish swims with wave amplitude 0.08 L, it shows higher efficiency than with the wave amplitude 0.1L, while it achieves the equal cruising velocity in the range of 0.30~0.65 L/s. The major frequency of pressure data is equal to the wave motion frequency. The magnitude of major frequency increases monotonically with wave amplitude and frequency, which presents approximate linear relationship. The flow adaptation control method based on pressure sensing showed satisfactory convergence and performance.The work in this paper explores the flow-sensing-in-loop adaptation control for bioinspired underwater robots. Moreover, the work is also full of significance to develop the more maneuverable and higher efficient underwater vehicles.