Application of ANFIS for machinery noise prediction in a bauxite mine

Abstract

The aim of this study is to predict the sound pressure level (SPL) values from ISO 9613-2, arti?cial neural network (ANN) and adaptive neuro fuzzy inference system (ANFIS) model. For this reason, sound pressure levels produced from the machinery operations were measured in a highly mechanized opencast bauxite mine in India. After collecting the ordered pairs of sound power level, distance, temperature, relative humidity and wind speed, SPL values were determined using ISO 9613-2 method. An endeavour has been made to examine the applicability of the developed soft computing models to predict SPL. To accomplish this objective, data obtained from the machinery operation measurements were assessed by constructing an ANFIS based prediction model. Sound power level (SWL), distance, temperature, relative humidity and wind speed were chosen as the input parameters and SPL as the output of the developed model. SPL prediction capability of the constructed ANFIS model has proved to be effective in terms of performance indices such as root mean square error (RMSE) between anticipated and measured SPL values. Prediction performance comparison has been made between the measured, ISO 9613-2 model, ANN and ANFIS model. The results indicated that the proposed ANFIS model exhibited better prediction performance when compared with ANN prediction model.