Depth control analysis of Smartfloat using sea trail data and parameter identification

Ping Cheng,Junjun Cao,Jialuan Xiao,Yiming Zhong,Caoyang Yu
DOI: https://doi.org/10.1016/j.oceaneng.2024.118391
IF: 5
2024-06-09
Ocean Engineering
Abstract:Accurate depth control and prediction are fundamental for scientific observations on time and space scales. In this study, the depth control performance of Smartfloat, which is a novel observation platform that integrates Argo and Glider, was analyzed by integrating sea-trail data and the adaptive step-size least squares (ASLS) parameter identification algorithm. First, a one-dimensional Smartfloat submersible model that considers the nonlinear characteristics of the buoyancy drive system was established by accounting for the influences of seawater density and pressure shell contraction. The discrete motion model was then reconstructed using the expression form of the correlation function, to reduce the influence of external interference on the identification. The anti-data saturation function of the parameter identification algorithm was realized by adjusting the value of the covariance matrix. Finally, sea trial data with a large depth and full profile were analyzed following a trial conducted in the South China Sea in 2022, and the theoretical calculation results were compared with the parameter identification and depth prediction results. The results revealed that when compared with traditional least squares and support vector machine identification algorithms, the proposed ASLS algorithm demonstrates a higher convergence speed and smaller prediction error. The ASLS algorithm can predict the performance parameters of large-delay drive systems such as underwater vehicles, and provides certain models and data support for the accurate control and prediction of depth.
engineering, civil, ocean, marine,oceanography
What problem does this paper attempt to address?