Izvorni znanstveni članak

Identification of Unmanned Underwater Vehicles by Self-OscillationMethod

Nikola Mišković orcid.org/0000-0003-1474-4126 ; Fakultet elektrotehnike i računarstva, Sveučilišta u Zagrebu, Zagreb, Hrvatska
Zoran Vukić orcid.org/0000-0002-5998-2913 ; Fakultet elektrotehnike i računarstva, Sveučilišta u Zagrebu, Zagreb, Hrvatska
Matko Barišić ; Fakultet elektrotehnike i računarstva, Sveučilišta u Zagrebu, Zagreb, Hrvatska

Sažetak

Control of underwater vehicles is a challenging task since these systems demonstrate highly coupled and nonlinear behavior in uncertain and often unknown environment. In order to successfully design higher levels of control hierarchy, sufficiently accurate parameters of a mathematical model describing the vessel is required. These parameters vary significantly depending on the payload; hence conventional, time-consuming identification methods are tedious. This paper introduces a self-oscillation based method for determining inertia and drag parameters for underwater vehicles. The procedure is easily implementable in field conditions and gives satisfactory results. Both linear and nonlinear models of yaw, heave and surge degree of freedom can be identified. Experimental results obtained from yaw identification experiments on a real underwater vehicle will be presented. In addition to this, the same methodology will be used to determine which model describes the vehicle dynamics more suitably. Modifications of the proposed algorithm for systems with delays and discrete-time systems will be described, together with an estimate of parameter error bounds due to quantization levels.

Ključne riječi

Nonlinear systems; Unmanned underwater vehicles; Identification; Self-oscillations

Hrčak ID:

47424

URI

https://hrcak.srce.hr/47424

Datum izdavanja:

22.12.2009.

Podaci na drugim jezicima: hrvatski

Posjeta: 1.801 *