Machine learning prediction of pyrolytic gas yield and compositions with feature reduction methods: Effects of pyrolysis conditions and biomass characteristics

This study aimed to utilize machine learning algorithems combined with feature reduction for predicting pyrolytic gas yield and compositions based on pyrolysis conditions and biomass characteristics. To this end, random forest (RF) and support vector machine (SVM) was introduced and compared. The results suggested that six features were adequate to accurately forecast (R-2 > 0.85, RMSE < 5.7%) the yield while the compositions only required three. Moreover, the profound information behind the models was extracted. The relative contribution of pyrolysis conditions was higher than that of biomass characteristics for yield (55%), CO2 (73%), and H-2 (81%), which was inverse for CO (12%) and CH4 (38%). Furthermore, partial dependence analysis quantified the effects of both reduced features and their interactions exerted on pyrolysis process. This study provided references for pyrolytic gas production and upgrading in a more convenient manner with fewer features and extended the knowledge into the biomass pyrolysis process.

Automated summary

This study aimed to predict pyrolytic gas yield and compositions based on pyrolysis conditions and biomass characteristics using machine learning algorithms and feature reduction. The results showed that six features were sufficient for accurate yield prediction, while compositions only required three. The study revealed the higher relative contribution of pyrolysis conditions to yield, CO2, and H2 compared to biomass characteristics.