20.500.12556/RUL-131666 Strojno določanje delovnih operacij pri sečnji z motorno žago s pomočjo ropota. Mechanical determination of work operations in felling with a chainsaw by means of rumble. Ker je terensko zbiranje podatkov zelo zamudno in drago opravilo, je bil namen diplomske naloge ugotoviti, kako natančno je mogoče določiti dele delovnega časa in delovne operacije sečnje z motorno žago s pomočjo podatkov o jakosti ropota in strojnega učenja. Poleg treh metod strojnega učenja (Naključni gozd, Nevronska mreža, Odločitveno drevo) smo določevanje izvedli tudi s pomočjo pogojnih stavkov v MS Excel programu. Podatke o sestavi delovnega časa in jakosti ropota smo razdelili na učno (27 dreves) in testno množico (26 dreves). Kot najuspešnejša metoda se je izkazala Naključni gozd. Efektivni in obratovalni čas motorne žage je bil določen s 96,0 % in 99,4 % natančnostjo, glavni in pomožni produktivni čas z 92,0 % in 65,7 % natančnostjo in podfazi podiranje in izdelava lesa pa s 84,1 % in 94,9 %. Največja natančnost določanja delovnih operacij je bila pri izdelavi zaseka in kleščenju (80,3 % in 89,5 %), najmanjša pa pri beljenju panja, obdelavi korenovca in prežagovanju (od 7,6 % do 14,0 %). Raziskava je pokazala na velik potencial strojnega učenja za pridobivanje podatkov o sestavi časa. As field data collection is a very time-consuming and costly task, the aim of the thesis was to determine how accurately parts of the working time and working operations of chainsaw felling can be determined using noise data and machine learning. In addition to the three machine learning methods (Random Forest, Neural Network, Decision Tree), we also performed the determination using conditional statements in MS Excel. The data on the composition of the working time and noise exposure were split into a training set (27 trees) and a test set (26 trees). Random Forest has proven to be the most successful method. The effective and operating time of the chainsaw was determined with 96.0% and 99.4% accuracy, the main and auxiliary productive time with 92.0% and 65.7% accuracy and the sub-phases of felling and timber production with 84.1% and 94.9% accuracy. The highest accuracy in determining work operations was observed in the case of notch-cutting and delimbing (80.3% and 89.5%, respectively), while the lowest accuracy was observed in the case of stump debarking, butt trimming and cross-cutting (7.6% to 14.0%). The results show great potential for collecting time study data through machine learning. strojno učenje sečnja ropot motorna žaga machine learning felling noise chain saw true false false [R. Petrovčič] Slovenski jezik Angleški jezik Diplomsko delo/naloga 2021-10-01 08:15:02 2021-10-01 08:15:06 2022-09-10 04:13:01 0000-00-00 00:00:00 2021 0 Ljubljana 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 630*30+630*36(043.2)=163.6 199188 78991363 21244.pdf 21244.pdf 1 F44BADEF6B8E2EE0705ED3B0276C1AFD 161c8f0f91bacfde5c17debc8e6e06eaf966e562dee37796e44473a5ac56fa91 a53b9141-227e-11ec-a523-00155dcfd717 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=149053 Biotehniška fakulteta 0 0 0