Acoustic Emission Techniques Combined with Machine Learning for Particle Size Distribution in Solid-Gas Fluidized Beds

The characterization of particles in solid-gas flows is of great importance in many industrial sectors such as nuclear energy and pharmaceuticals industries. Acoustic emission (AE) techniques can provide non-intrusive, multi-point measurements in non-transparent test sections and complement other techniques e.g., optical sensors, x-ray imaging. The combination of AE techniques with machine learning algorithms (ML) is particularly beneficial for industrial settings, because it reduces the cost of signal post processing and is suitable for both small and large-scale experiments. In this paper, we develop an acoustic emission technique in combination with machine learning algorithms to characterize the particle size distribution in a solid-gas fluidized bed. The theory underpinning the generation of an acoustic emission signal in solid-gas flows is explained. An AE signal is generated in gas-solid fluidized beds in the form of an elastic wave with frequencies >100 KHz and it propagates through the gas-solid mixture. An inversion algorithm to extract the information on the particle size from the energy of the AE signal is described. Experiments were carried out in a pseudo-2D flat fluidized bed with four glass bead samples with different sizes, ranging from 100 µm to 710 µm. AE signals were recorded with sampling frequency of 5 MHz. The AE signal post processing and the data preparation for the ML process are explained. For the ML process, the AE frequency, AE energy and particle velocity data were divided in training (60%), cross validation (20%) and test (20%) sets. Two ensemble ML approaches, namely Random Forest and Gradient Boosting are applied to predict particle sizes based on the AE signal features with an accuracy greater than 99.5% and a R squared value larger than 0.999. Experimental results indicate that the AE technique is a powerful tool for the characterization of the particle size in solid-gas flows.