Predicting biomass comminution: Physical experiment, population balance model, and deep learning

• DOI: 10.1016/j.powtec.2024.119830
• Publication Date: 2024-05-04T09:33:48Z
• Authors: Minglei Lu; Yidong Xia; Tiasha Bhattacharjee; Jordan Klinger; Zhen Li
• Keywords: Comminution

An extended population balance model (PBM) and a deep learning-based enhanced neural operator (DNO+) are introduced for predicting particle size distribution (PSD) of comminuted biomass through large knife mill. Experimental tests using corn stalks with varied moisture contents, mill blade speeds, discharge screen sizes conducted to support development. A novel mechanism in the PBM allows including additional input parameters such as content, which is not possible original PBM. The DNO+ can include influencing factors different data types content size, significantly extends engineering applicability standard DNO that only admits feed PSD outcome PSD. Test results show both models remarkably accurate calibration or training parameter space be used surrogate provide effective guidance preprocessing design.