o:3031 Klasifikacija motora automobila sa unutrašnjim sagorevanjem prema pogonskom gorivu na osnovu generisanog zvuka Classification of internal combustion vehicle engines according to fuel based on generated sound : doctoral dissertation sr Noise pollution and exhaust emissions represent significantenvironmental challenges, with vehicle classification based on the type offuel contributing to the reduction of air pollution and improvement of urbanspace management. This research develops an original system for acquiringengine sounds of passenger vehicles under real-world conditions, enablingautonomous vehicle detection, minimal environmental impact, andcompatibility with future IoT systems. Key challenges included optimizingthe microphone position to avoid obstructing vehicle passage, as well asselecting the engine operating mode that carries the most relevant acousticinformation. The system was tested on 350 vehicles, forming a database of475 sound samples, along with the implementation of a method forautomatic extraction of acoustically relevant operating modes.Two groups of audio features suitable for distinguishing fuel typeswere identified: spectrogram representations (mel-spectrograms,gammatonegrams) and psychoacoustic characteristics (loudness, roughness,sharpness, crest factor). Classification was carried out using deep learningmethods (CNN) and unsupervised learning (SOM), with CNN achieving an F1score of up to 97%, and SOM reaching 96.7% after eliminating technicallyfaulty vehicles.The contribution of this research lies in the development of aninnovative acquisition system, the identification of key acoustic features,and the demonstration of machine and deep learning applications in enginesound analysis. The created sound sample database represents a valuableresource for the scientific community and future research in noise analysis,fuel type recognition, and fault detection. The results confirmed that it ispossible to successfully classify vehicles based on fuel type using sound,opening opportunities for traffic monitoring, noise reduction, and thedevelopment of standards for assessing engine sound quality. Biografija autora: list. 137Bibliografija: list. 124-136 Datum odbrane: 7.8.2025. Acoustics 91552100 177427465 91552101 8788 2025-10-21T13:22:43.053Z 45 no 46 Marko, 1986- Milivojčević 24438631 2025 63 mentor Dejan G., 1967- Ćirić 11375207 2025 63 član komisije Zoran, 1964- Perić 7648359 2025 63 član komisije Jelena, 1978- Nikolić 7035751 2025 63 član komisije Jelena D., 1970- Ćertić 27607399 2025 63 član komisije Milan, 1983- Dinčić 106116105 2025 VI, 137 listova 6841071 http://phaidrani.ni.ac.rs/o:3031 no yes 12 2025-10-21T13:22:43.320Z 70 Klasifikacija vozila, motori sa unutrašnjim sagorevanjem, analiza zvuka,akvizicija akustičkih signala, spektralna analiza, psihoakustička obeležja,duboko učenje, mašinsko učenje Vehicle classification, Internal combustion engines, Sound analysis,Acoustic signal acquisition, Spectral analysis, Psychoacoustic features, Deeplearning, Machine learning (534.87+534.836)004.85 T 121 1738 18A03 18A0302 2025