Inteligentno upravljanje brodskim procesima

scientific master's theses

• Author: Jasmin Velagić
• Mentor: Zoran Vukić (mentor)
• Committee member: Ljubomir Kuljača (predsjednik povjerenstva)
• Committee member: Zoran Vukić (član povjerenstva)
• Committee member: Vladimir Koroman (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Control and Computer Engineering); Zagreb
• Defense date and country: 1999-07-14, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Universal decimal classification (UDC): 004 - Computer science and technology. Computing. Data processing

Abstract

Cilj ovog rada je primjena inteligentnog upravljanja kod problema vođenja broda po kursu i trajektoriji. Opisani su najčešće korišteni postupci inteligentnog upravljanja i razlozi koji su doveli do njihove izgradnje. Kao autopiloti korišteni su neizraziti (Mamdani, Sugeno) i neuro autopiloti (BP, RB). Opisan je kompletan postupak sinteze neizrazitih autopilota. Neuro autopiloti uči od neizrazitog autopilota, pri čemu se zahtijeva da upravljačke površine neizrazitog i neuro autopilota budu identične (što je moguće bolje). Analiza problema vođenja broda po kursu i trajektoriji sprovedena je na nelinearnim modelima brodova Mariner i ESSO Tanker. Dinamičko ponašanje neizrazitih i neuro autopilota uspoređeno je sa konvencionalnim PID autopilotom. Analiziran je utjecaj valnih poremećaja na dinamičko ponašanje broda pri promjeni kursa i postupci za minimiziranje tog utjecaja. Kod problema vođenja broda po trajektoriji prikazan je manevar zaobilaženja putne točke i vođenje broda po složenoj putanji (trajektoriji). Utjecaj vanjskih hidrodinamičkih poremečaja uslijed mimoilaženja s drugim brodom analiziran je na modelu broda Mariner klase. Efekt plitke vode pri velikom manevru analiziran je na modelu broda ESSO Tanker. Također je razmatran utjecaj morske struje i valova na performanse sustava vođenja broda po trajektoriji. U tu svrhu su projektirani 3-ulazni neizraziti autopiloti (Mamdani, Sugeno), kako bi se poboljšale performanse sustava vođenja. Osim toga, projektiran je i inteligentni autopilot čiju osnovu čine neizraziti autopilot (Mamdani) i mehanizam podešavanja faktora skaliranja. Mehanizam podešavanja predstavlja neizraziti regulator koji mijenja faktore skaliranja osnovnog neizrazitog autopilota.

Abstract (english)

The aim of this paper is intelligent control implementation at course-keeping and track-keeping of ship. The most used intelligent control procedures are described, as well as the reasons that caused their development. Fuzzy logic controllers (Mamdani, Sugeno) and neural networks (Backpropagation, Radial Basis) are used as autopilots. The complete design procedure of fuzzy autopilots is described. Neural autopilot learns from fuzzy autopilot where it is demanded that control surfaces of fuzzy and neural autopilots be identical (as better as possible). Analysis of ship course-keeping and track-keeping problems was performed on the nonlinear ship models Mariner i ESSO Tanker. Dynamical behavior of fuzzy and neural autopilots is compared with conventional PID autopilot. The influence of wave disturbance on dynamical behavior of ship during course-changing and methods for minimizing of that influence are analyzed. On the track-keeping problem the maneuver of way point turning and ship guiding through complex path (trajectory) are presented. The influence of the external hydrodynamic disturbance from passing ship is analyzed on Mariner Class Vessel. The shallow water effect during large maneuver is analyzed on ESSO Tanker. The influence of sea current and wave disturbances on track-keeping performance was also considered. In that purpose 3-input fuzzy autopilots (Mamdani, Sugeno) are designed in order to improve control system performance. Besides, intelligent autopilot is designed and it consists of Mamdani type fuzzy autopilot and mechanism for adjusting of scaling factors. The adjustment mechanism represents fuzzy controller that changes scaling factors of the base fuzzy autopilot.

• Keywords: Inteligentno upravljanje; autonomni sustav; neizraziti regulator; neuronska mreža; planiranje; detekcija kvarova; vođenje broda po kursu; vođenje broda po trajektoriji; poremećaji; inteligentni autopilot; adaptivni neizraziti autopilot
• Keywords (english): Intelligent control; autonomous system; fuzzy controller; neural network; planning; fault detection; course-keeping; track-keeping; disturbances; intelligent autopilot; adaptive fuzzy autopilot
• Language: croatian
• URN:NBN: urn:nbn:hr:168:521521
• Study programme: Title: Postgraduate master programe in computing; Study programme type: university; Study level: postgraduate; Academic / professional title: magistar inženjer računarstva (magistar inženjer računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-03-20 13:21:49