A study on the enzymatic hydrolysis of steam exploded napiergrass with alkaline treatment using artificial neural networks and regression analysis

摘要: Abstract To facilitate the enzymatic saccharification of widely available lignocellulosic biomass, napiergrass was subjected to a two-stage pretreatment process consisting steam explosion (SE) followed by alkaline delignification. SE performed under various reaction temperatures, times, and particle sizes. The experimental results show that effective at removing xylan lignin significantly enhanced digestibility napiergrass. Up 85% in original material removed, leaving cellulose-rich residue highly susceptible hydrolysis; reached 96.1%. alone not sufficient, because considerable amount still remained steam-exploded solids. In addition work, models predicting as function conditions were developed using artificial neural network regression techniques, their performance compared. Three different methods used, i.e. , back-propagation (BPNN), multiple linear (MLR), partial least-square (PLS). input model three parameters (temperature, time, size), while output. BPNN provided reasonable predictive performance. addition, temperature found be most significant factor among studied, time.