Optimizing Biomass Pyrolysis: A Comparative Analysis of GA, PSO, and SCE Algorithms

Renewable energy, especially biomass pyrolysis, are receiving increasing attention due to their economic and environmental benefits. To advance biomass pyrolysis technologies, it is crucial to accurately determine the kinetic parameters, with heuristic algorithms like Genetic Algorithm (GA), Hill Climbing (HC), Particle Swarm Optimization (PSO), and Shuffled Complex Evolution (SCE) proving more efficient than traditional methods. However, there’s a lack of comparative studies on these algorithms’ effectiveness in optimizing biomass pyrolysis kinetics.

In 08 September 2023, Emergency ManagementScience and Technology published a research article entitled by “A comparative study of GA, PSO and SCE algorithms for estimating kinetics of biomass pyrolysis”.

This study utilized thermogravimetric analysis to explore the kinetics of wood pyrolysis at heating rates of 5, 10, and 20 K/min, distinguishing between the different phases of wood decomposition through the Gauss multi-peak fitting method for more precise kinetic parameter estimation. The efficiency and accuracy of three optimization algorithms, namely Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Shuffled Complex Evolution (SCE) were then assessed for optimizing these kinetic parameters. The experimental results highlighted how higher heating rates shift decomposition reactions to higher temperatures, with wood pyrolysis involving water evaporation, and the decomposition of hemicellulose, cellulose, and lignin. Through analytical methods, including the Kissinger and KAS methods, activation energy and frequency factors were derived, demonstrating minor deviations across methods, suggesting all the reactions could be considered as single-step due to the small variation in activation energy.

Comparative analysis of the optimization algorithms revealed SCE’s superior accuracy but lower computational efficiency compared to GA and PSO, which displayed comparable computational efficiencies. PSO is the most well-balanced algorithm with high accuracy, computational efficiency and convergence efficiency, but its tendency to search for local optimums is considered as a potential drawback. The validation of these algorithms through numerical prediction of experimental data confirmed their reliability, with some deviations pointing the complex interactions between kinetic parameters in the optimisation process.

In conclusion, although no one algorithm outperformed the others in all metrics, PSO achieved a good balance between accuracy, efficiency and speed. This suggests that future research direction towards developing hybrid or AI-enhanced algorithms that could leverage the strengths of each to optimize the complex kinetics of biomass pyrolysis more effectively.

###

References

DOI

10.48130/EMST-2023-0009

Original Source URL

https://www.maxapress.com/article/doi/10.48130/EMST-2023-0009

Authors

Hongfang Wang, Junhui Gong^*

Affiliations

College of Safety Science and Engineering, Nanjing Tech University, Nanjing, Jiangsu 211816, China

About Junhui Gong

Professor, College of Safety Science and Engineering, Nanjing Tech University. He has long been engaged in research on solid pyrolysis ignition, fire spread, thermal runaway of lithium batteries, hydrogen leakage diffusion, and combustion processes of composite and flame retardant materials.