Recursive least square algorithm and sliding mode controller to diving system of an AUV

The problems of model identification and depth control for an autonomous underwater vehicle (AUV) are addressed in this paper. In order to enhance the controller´s reliability, estimation of parameters using experimental data from previous lake tests of an AUV named BSA is carried out to minimize modeling error. With the identified hydrodynamic model in the vertical plane, this paper adopts an optimal sliding mode control scheme to solve the depth control problem in the presence of vehicle´s nonlinear terms and model uncertainty. This proposed method verified by simulation and lake trial shows that it is effective to realize depth control with the help of system identification.