Assessing acoustic emission in 1055I John Deere combine harvester using statistical and artificial intelligence methods

Abstract

Agricultural mechanisation is accompanied by several challenges, one of them is the noise pollution caused by machinery. Noise pollution has undesirable effects on humans such as temporary or permanent loss of hearing, decrease in working efficiency and increase in accidents. The aim of this study is to assess noise pollution in the 1055I John Deere combine harvester. The tests were performed at different engine speeds, gear positions and sound measuring locations. The obtained data were analysed in the form of factorial test based on a completely randomised design. An artificial neural network model was created to predict the sound level in the combine. The results of variance analysis for the effects of the main factors on the level of sound were significant at probability level of 1%. The sound intensity reaching the driver's ear based on frequency analysis is equal to 84.16 dB at the frequency of 4000 Hz. The mean square error and the correlation coefficient for the best neural network with ten neurons in the hidden layer were obtained. The permitted time duration of driving the harvester was calculated to be less than 2 h.