Determination of the heat capacity of cellulosic biosamples employing diverse machine learning approaches

• Published in: Energy science & engineering 2022-06, Vol.10 (6), p.1925-1939
• Main Authors: Karimi, Mohsen, Khosravi, Marzieh, Fathollahi, Reza, Khandakar, Amith, Vaferi, Behzad
• Format: Article
• Language: English
• Subjects: Accuracy; Agricultural production; Alternative energy; Artificial intelligence; Biodiesel fuels; Biofuels; Biomass; Carbon dioxide; Cellulose; cellulosic sample; computational modeling; Data analysis; Empirical analysis; Fossil fuels; Heat; heat capacity; Kernel functions; least‐squares support vector regression; Lignin; Lignocellulose; Machine learning; Network topologies; Neural networks; Outliers (statistics); Specific heat; Sulfur; Support vector machines; Sustainable development; Thermal properties; Thermodynamic properties
• ISSN: 2050-0505; 2050-0505
• DOI: 10.1002/ese3.1155

Heat capacity is among the most well‐known thermal properties of cellulosic biomass samples. This study assembles a general machine learning model to estimate the heat capacity of the cellulosic biomass samples with different origins. Combining the uncertainty and ranking analyses over 819 artificial intelligence models from seven different categories confirmed that the least‐squares support vector regression (LSSVR) with the Gaussian kernel function is the best estimator. This model is validated using 700 laboratory heat capacities of four cellulosic biomass samples in wide temperature ranges (absolute average relative deviation = 0.32%, mean square errors = 1.88 × 10−3, and R2 = 0.999991). The data validity investigation approved that only one out of 700 experimental data is an outlier. The LSSVR model considers the effect of the cellulosic samples' crystallinity, temperature, and sulfur and ash content on their heat capacity. The overall prediction accuracy of the LSSVR is more than 62% better than the achieved accuracy using the empirical correlation. Research highlights Heat capacity of the cellulosic biosamples with different origins is simulated. Seven different intelligent estimators have been utilized for the modeling stage. Least‐squares support vector regression shows the most accurate predictions. This approach has the overall absolute average relative deviation (AARD) = 0.32%, mean square errors (MSE) = 1.88 × 10−3, and R2 = 0.999991. The amorphous cellulosic biosample has the highest average heat capacity.