One of the most important ergonomic and industrial issues is the human health when they are
working with machines in particular agricultural machines, which was equipped with diesel engines.
Users of agricultural machines are exposed to high level of vibration and working more
than allowable time causes damaging different part of body, decreasing efficiency and quality of
work. Nowadays, the most of fossil fuels is consumed by the combustion engines. Combustion
of fossil fuels on the engines causes rapid depletion of energy sources and also harmful effect on
environment. A suitable alternative fuel for diesel engines is called biodiesel due to their vital
source. The vibration characteristics of diesel engine could be changed when using blended diesel-biodiesel
fuel. In this research, the allowable vibration exposure time was estimated with
ANFIS methodology for two-wheel tractor fuelled by diesel-biodiesel blends. A 13-hp twowheel
tractor with 5 types of fuels was used. The tractor was driven with five engine speeds and
four gear ratios. To gather the vibration acceleration data of this tractor, three accelerometers
were installed in three mutually perpendicular directions on seat of trailer pulled by tractor. The
results indicated that the best fuzzy inference system structure was subtractive clustering. The
optimum radius of subtractive clustering was 0.75, 0.65 and 0.80 As well; the number of rules
was 10, 18 and 9 for longitudinal, lateral and vertical directions, respectively. In all of the three
directions, the differences between predicted and measured values are decreased with increasing
the frequencies 1 to 80 Hz. The developed ANFIS network could be predicted the vibration acceleration
of combustion frequencies and also the trend of 1/3rd octave frequency spectra. The
coefficient correlations of model within frequency range of 1-80 Hz were 0.75, 0.85 and 0.91 for
the longitudinal, lateral and vertical directions, respectively.