Estimation of hydrodynamic coefficients and simplification of the depth model of an AUV using CFD and sensitivity analysis

Abstract

Maneuverability is one of the most important performance characteristics of submarines. Before constructing a designed AUV, the hydrodynamic model can be used to determine its inherent motion behavior. Standard submarine motion equations (SSME) are the most widely used hydrodynamic model in which more than 100 coefficients must be estimated. These coefficients are usually determined through conventional experimental and analytical methods. The most common method today is captive model tests, but the separately determining approach is time-consuming and makes it difficult to assess the reliability of the model. This paper proposes an efficient approach for estimating hydrodynamic coefficients (HCs) using computational fluid dynamics (CFD). Instead of captive model tests, the proposed virtual free-running test can provide all the information required to determine all the HCs in only one simulation. Kalman filter estimation methods are used to determine the HCs. Sensitivity analysis and statistical results show that in a typical AUV maneuvering, SSME can be simplified. The final simplified equations of motion have much fewer components than the original model while fitting accuracy remains. Using the experimental data of the well-known DARPA-SUBOFF underwater vehicle, it is shown the proposed virtual free-running and the simplification approaches are effective and reliable.