Robustness of autonomous underwater vehicle control in variable working conditions

vehicles. Sliding mode control was one of the fi rst control methods implemented for underwater vehicles. It has shown advantages in several implementations. Different control methodologies used for underwater vehicles are categorized by Fossen. In the work of Logan, H/msynthesis methods are compared with the sliding mode method. A comparison of different control techniques, including artifi cial intelligence methods, is presented by Lea et al. While H and m-synthesis methods can handle the discarded higher-order modes of a system, they may lead to conservative results in the computation of allowable variations in the parameters of the system to maintain stability, since the structure of the parameter uncertainty is not taken into account in these methods. Considerable research effort has been invested in control of systems with uncertain parameters by using the properties of their characteristic polynomials. In work by Ackermann, the applicability of polynomial methods to the steering of ground-based and fl ying vehicles was demonstrated. When the characteristic polynomials of linear systems are linear functions of uncertain parameters, there exists a wide range of tools to deal with such systems. However, when the coeffi cients of the characteristic polynomials are nonlinear or multilinear functions of the parameters, only few results are available. Several works on this subject are gathered in the books by Ackermann and Djaferis. In this article, an underwater vehicle, whose operating speed and hydrodynamic derivatives are subject to perturbations, was modeled as a plant with uncertain parameters. The vehicle under study is Naval Postgraduate School (NPS) Autonomous Underwater Vehicle (AUV) II that has been described by Healey and Leinard. The effects of the uncertainty in the operating velocity Abstract The performance of the control systems of autonomous underwater vehicles (AUVs) in the presence of parameter variations was studied. With an AUV working at different operating speeds and in different ocean environments, the physical parameters such as speed, hydrodynamic coeffi cients, or inertias may be perturbed from their nominal values. The vehicle control systems can be modeled as systems with parameter uncertainty. An existing robust control method, which uses the robustness properties of polynomials, was used for this system to calculate the permissible ranges of variation in the parameters. The method was applied to the Naval Postgraduate School AUV II and the results were verifi ed by simulating the motion control of the vehicle under the infl uence of parameter perturbations.