Predicting the high heating value and nitrogen content of torrefied biomass using a support vector machine optimized...

Area of Science:

• Biomass energy conversion
• Computational modeling
• Sustainable fuels

Background:

• Torrefied biomass is a promising solid fuel alternative.
• Accurate prediction of torrefied biomass properties (HHV, No) is crucial for its effective utilization.
• Existing prediction models may lack precision for feedstock properties and torrefaction conditions.

Purpose of the Study:

• To develop and optimize a predictive model for the higher heating value (HHV) and nitrogen content (No) of torrefied biomass.
• To enhance the predictive accuracy of a support vector machine (SVM) model using sparrow search algorithm (SSA) optimization.
• To validate the model's performance against experimental data for reliable biomass fuel characterization.

Main Methods:

• Development of a support vector machine (SVM) model incorporating a Radial Basis Function (RBF) kernel.
• Optimization of the SVM model parameters using the sparrow search algorithm (SSA).
• Prediction of higher heating value (HHV) and nitrogen content (No) based on feedstock properties and torrefaction conditions.

Main Results:

• The sparrow search algorithm (SSA) significantly improved the prediction performance of the support vector machine (SVM) model.
• The optimized SSA-SVM model achieved a coefficient of determination (R²) greater than 0.91 for both HHV and No predictions.
• Root Mean Square Error (RMSE) values were acceptable, and predicted values closely agreed with experimental data, demonstrating high predictive precision.

Conclusions:

• The SSA-optimized SVM model offers a highly precise and reliable method for predicting the HHV and No of torrefied biomass.
• This predictive capability serves as a valuable reference for the efficient utilization of torrefied biomass in solid fuels.
• The study provides insights for the optimized design of torrefaction facilities and processes.