Artificial neural network (ANN) is one kind of artificial intelligence in the computing system that aims to process information as the way neurons in the human brain. In this study, the combination of the Taguchi method and ANN are used to maximize and predict biofuel yield from spent mushroom substrate torrefaction and pyrolysis via microwave irradiation. The Taguchi method is utilized to design the multiple factors (particle size, catalyst, power, and magnetic agent) and levels of experiment parameters. The highest total biofuel yield (biochar + bio-oil) is 99.42%, accomplished by a combination of 355 µm particle size, 300 mg·g-SMS^-1 catalyst, 900 W power, and 300 mg·g-SMS^-1 magnetic agent. ANN with one hidden layer shows the outstanding linear regression predictions for the highest biofuel yields (biochar 0.9999 and bio-oil 0.9998). This high linear regression indicates that ANN with a quick propagation algorithm is an appropriate approach for predicting biofuel conversion.

人工神经网络（ANN）是计算系统中的一种人工智能，其目的是像人脑中的神经元一样处理信息。在这项研究中，田口方法和人工神经网络的组合被用来最大化和预测来自用过的蘑菇基质烘焙和微波辐射热解的生物燃料产量。田口方法用于设计多个因素（粒度、催化剂、功率和磁性剂）和实验参数的水平。最高总生物燃料产率（生物炭 + 生物油）为 99.42%，由 355 µm 粒径、300 mg·g-SMS ^-1催化剂、900 W 功率和 300 mg·g-SMS ^{-1 的组合实现}磁性剂。具有一个隐藏层的 ANN 显示了对最高生物燃料产量（生物炭 0.9999 和生物油 0.9998）的出色线性回归预测。这种高线性回归表明具有快速传播算法的 ANN 是预测生物燃料转化的合适方法。