| Autonomous underwater vehicle (AUV) is a typical complex engineering systeminvolving many coupling disciplines. Designing an AUV should deal with hull form, pressureshell, frame structure, energy selection, propulsion, manoeuvre and control mode and so on.The conventional design method for AUV is carried through orderly following the principleof different phases and different disciplines, which results in the reiteration in the designcourse. Besides, the optimal result of AUV system is often lost because of an absence ofbalanced consideration among disciplines. Consequently, the design efficiency and systemperformance are decreased using conventional design method.Multidisciplinary design optimization (MDO) which can well exploit the mutuallyinteracting phenomena of different subsystems provides an effective method for optimaldesign of complex system. Aimming at figuring out how to compromise and achieving datatransfer among disciplines during the optimal problem of AUV general design, an AUVwhose job depth is2000m and working range is100km is the objective in this study. TheMDO theory and its application in AUV system are studied for two structural forms: themainstream frame structure made of metal and novel integral structure made of carbon fiber.The main work is as follows:(1) Six aspects are considered in general design of AUV: hull form, hull frame, energyselsction, pressure shell, propulsion and general layout. The methods of analysis andestablishment of mathematic models for relevant disciplines are summarized. The mutuallycoupling characteristic among subsystems is studied, and the design variables, objectives andconstraints of subsystems and system are defined. Finally, the relations between system andsubsystems are explicated, also the relations among subsystems are explicated.(2) Approximation model which is one of the most important technologies in the MDOtheory is studied. Based on design of experiments and approximation technology, finiteelement results of serial spherical pressure shell and cylinder pressure shell stiffened by ringsare gained, and different approximation models of stresses and inflective pressure areestablished according these results. Approximation models of drag for serial revolution shapeare constructed using these samples and responses obtained by hydrodynamic numericalcalculation, and the precisions of these models are tested by flume experiments of serialmodels. (3) Two typical MDO methods that are simultaneous analysis and design andcollaborative optimization are studied in this paper, and the calculational specialty andadaptive field of the two methods are analysed according a mathematical example. Thearithmetic of NSGA-II is investigated to deal with multi-objectives problem inmultidisplinary optimization. MDO is applied to two typical problems in AUV design: theintegrative design of hull form and energy and the integrated design of resistance andmanoeuvre.(4) The scheme of metal frame structure for object AUV is designed, and fivesubsystems that are energy, hull form, manoeuvre, pressure shell and frame structure are setup.The inputs and outputs of all of the five mathematical model for subsystems and theoptimal objective,constraints and variables for system are defined. According to thesimultaneous analysis and design method, five subsystems are integrated. Finally, the optimalresult that can satisfy the assignment book is gained after MDO calculation.(5) The scheme of integral structure made of carbon fiber for object AUV is designed,the structural strength analysis is researched and APDL command lines of strength analysisbased on parameters are set up. Four subsystems that are energy, hull form, manoeuvre andframe structure are integrated. Also, the optimal result that can satisfy the assignment book issuccessfully gained after MDO calculation.Finally, the scheme of structure made of carbonfiber which has a big advantange over that made of metal is adopted by the objective AUV.It turns out that the MDO process of AUV general design in this paper covers manyprimary contents and the optimal result can give information of many important parameters ofenergy capability, pressure structure, carrier structure, rudder area and so on. The optimalresult can be efficiently gained from the MDO process only by setting the design variable andcontrol conditions without man-made reiteration in the design course. The method haspreferable practicability and universal property for AUV design. |
PDF Full Text Request