Designing an Intelligent Expert Control System Using Acoustic Signature for Grinding Mill Operation

This paper presents an instrumentation scheme to use acoustic signal as the control signal to regulate the operation of a grinding mill. Due to the very operational nature of the grinding mill, it is not hard rather impossible to install conventional sensors like optical, temperature etc. inside the grinding mill for obtaining feedback signal which could monitor the grinding operation. Therefore, conventional functioning of the grinding mill does not provide a mechanism to achieve optimum grinding by regulating the parameters of the mill in case either mill enters into an erroneous state or the product deviates from the desired range. Moreover, on study it has found that there is no scheme existing to predict grinding operation by which parameters of the mill could be tuned. In this work an appropriate theoretical background has also been established to predict dynamic breakage characteristics with respect to particle size distribution of materials, adequately supported by experimental data. The breakage characteristic has been verified using a neural network model while the validation of the learning scheme is also demonstrated in the paper.