Přístupnostní navigace
E-application
Search Search Close
Master's Thesis
Author of thesis: Bc. Martin Malatinec
Acad. year: 2025/2026
Supervisor: Ing. Ondrej Mihálik
Reviewer: doc. Ing. Miroslav Jirgl, Ph.D.
This master's thesis addresses the problem of identification and modeling of the dynamic properties of a human operator in a human-machine system during vehicle control. The main objective is to develop a software framework in the Matlab environment for automated processing of data obtained from a vehicle simulator and to compare the efficiency of different driver model structures. The theoretical part presents a review of modern approaches to human factor modeling, where the driver is characterized by classical linear models of the McRuer and Donges type, as well as by nonlinear systems represented by fuzzy systems and recurrent neural networks. The practical part comprises data collection from 15 respondents in four driving scenarios differing in the character of the disturbance signal, and the implementation of two model structures: a Mamdani fuzzy PD controller optimized by a genetic algorithm, and an LSTM recurrent neural network. Significant attention is devoted to a cross-validation, which verifies the ability of the models to generalize the learned control strategy to unseen data. The results discuss the capability of individual model structures to faithfully represent driver dynamics, highlighting the substantially better generalization properties of the LSTM network and the static Donges model.
dynamic driver models, human operator, fuzzy systems, LSTM neural networks, McRuer model, Donges model, cross-validation, vehicle simulator, human-machine systems
Date of defence
09.06.2026
Result of the defence
Defended (thesis was successfully defended)
Grading
D
Process of defence
Student odprezentoval a odpověděl na otázky oponenta. Student přiznal některé nedostatky, na které komise poukázala. Z výsledné obhajoby vyplývá, že práce postrádá matematický rigor, používá vágní pojmy a nesprávné nastavení algoritmů pro některé případy. Odpovědi studenta potvrzují neadekvátně provedenou literární rešerši.
Language of thesis
Slovak
Faculty
Fakulta elektrotechniky a komunikačních technologií
Department
Department of Control and Instrumentation
Study programme
Cybernetics, Control and Measurements (MPC-KAM)
Composition of Committee
prof. Ing. Michal Prauzek, Ph.D. (předseda) doc. Ing. Petr Beneš, Ph.D. (místopředseda) doc. Ing. Jakub Arm, Ph.D. (člen) Ing. Martin Čala, Ph.D. (člen) Ing. Jaroslav Lepka, Ph.D. (člen) prof. Ing. Radomil Matoušek, Ph.D. (člen) Ing. Lukáš Pohl, Ph.D. (člen)
Supervisor’s reportIng. Ondrej Mihálik
Grade proposed by supervisor: C
Reviewer’s reportdoc. Ing. Miroslav Jirgl, Ph.D.
Grade proposed by reviewer: E
Responsibility: Mgr. et Mgr. Hana Odstrčilová