Roof fall is one of the most important problems
connected with underground coal mines because it plays a
significant role in financial and human losses. Hence, it is
essential to accurately predict the roof fall rate for the
purpose of controlling, reducing, and/or even eliminating
the risk of the involved problems. On the other hand, there
are many different parameters that make a considerable
impact on the occurring roof rate. Most of these factors are
not completely known or measurable. Therefore, the
problem of predicting roof fall is vague, sophisticated, and
complex. Adaptive neuro-fuzzy inference system (ANFIS)
is a powerful and robust tool for modeling linear and nonlinear
patterns in science and engineering problems. In this
paper, the ANFIS system is applied to model the roof fall
rate in coal mines. The constructed model uses the subtractive
clustering method to generate fuzzy rules based on
data of roof performance from US coal mines. The
results demonstrate that prediction of roof fall rate by the
ANFIS model is satisfactory
