Bionic Parameter Extraction for Robotic Fish Design: A Reverse-Engineering Approach Using Large Yellow Croaker

Abstract

The bionic research on fish swimming is the theoretical basis for the design of robotic fish. Accurately extracting the parameters of the fish body wave equation is the key to achieving efficient biomimetic motion. This study adopted an innovative reverse parameter extraction method to accurately quantify the hydrodynamic characteristics of fish swimming based on real biological fish motion data. Five adult yellow croakers (Larimichthys crocea) were selected as experimental subjects. Continuous video capture was conducted for 7 hours at a sampling rate of 30 fps in a constant-temperature two-dimensional water environment. By utilizing OpenCV digital image processing technology, an automatic extraction workflow was established to convert the original video into the centerline of the fish body, including key steps such as grayscale conversion and skeleton extraction. On this basis, using the Lighthill slender body theory framework, the key parameters in the fish wave equation were accurately determined through point marking, coordinate transformation, and discrete fitting methods. The experimental results show that the tail beat period of the large yellow croaker is 0.5s (frequency f=2Hz), the maximum amplitude of tail beat A is 3.25cm, and the amplitude to body length ratio is 2A/L ≈ 0.19, which conforms to biological laws; The average swimming speed U is 13.3cm/s, with a slip ratio of 0.86, which is in line with theoretical expectations; The parameters of the fish body wave equation obtained through envelope fitting are C ₁=-0.075, C ₂=0.5, k=10, and ω=4 π. The waveform reconstructed based on these parameters in MATLAB has high similarity with the actual swimming motion of the large yellow croaker, verifying the accuracy of the fitting. The fish body wave parameters obtained by this method lay a reliable biological foundation for the subsequent precise design of the robotic yellow croaker.