Sensitivity analysis of control parameters errors and current parameters to motion accuracy of underwater glider using Sobol’ method

Abstract

Control parameters errors and current can weaken the motion accuracy of underwater gliders, which will obstruct the achievement of high accuracy exploration missions. To improve the glider motion accuracy more pertinently, it is necessary to identify the key factors that affect the glider motion accuracy. To resolve the above problem, we employ Sobol’ method to analyze the sensitivity of control parameters errors and current parameters to the glider motion accuracy. The control parameters errors include the net buoyancy adjustment amount error, the movable mass block translation amount error and the movable mass block rotation amount error. The current parameters include the current intensity, action depth and direction. First, the dynamic model of an underwater glider is established, and this model considers the effects of hull deformation, water density variation and current simultaneously. Then the dynamic model is validated by experimental data. After the evaluation parameter of motion accuracy is given, the glide motion accuracy is preliminarily analyzed, and then the sensitivity analysis is carried out. In this paper, we calculate the sensitivity coefficients by Monte Carlo method. To reconcile calculation efficiency with calculation accuracy, a convergence rule is proposed to reduce the number of sample points that are used for calculating sensitivity coefficients as much as possible. Besides, based on simulation results of dynamic model, the surrogate models are established to participate in the calculation of sensitivity coefficients, which can improve analysis efficiency further. For two typical motion modes, analysis results identify the key factors affecting the glider motion accuracy and illustrate the form of their effect on the glider motion accuracy. This research may be valuable for the achievement of the glider precision operation.