Catalytic thermal degradation of Chlorella vulgaris: Evolving deep neural networks for optimization

Sin Yong Teng; Adrian Chun Minh Loy; Wei Dong Leong; Bing Shen How; Bridgid Lai Fui Chin; Vítězslav Máša

doi:10.1016/j.biortech.2019.121971

Catalytic thermal degradation of Chlorella vulgaris: Evolving deep neural networks for optimization

Bioresour Technol. 2019 Nov:292:121971. doi: 10.1016/j.biortech.2019.121971. Epub 2019 Aug 8.

Authors

Sin Yong Teng¹, Adrian Chun Minh Loy², Wei Dong Leong³, Bing Shen How⁴, Bridgid Lai Fui Chin⁵, Vítězslav Máša¹

Affiliations

¹ Brno University of Technology, Institute of Process Engineering & NETME Centre, Technicka 2896/2, 616 69 Brno, Czech Republic.
² National HiCoE Thermochemical Conversion of Biomass, Centre for Biofuel and Biochemical Research, Institute of Sustainable Building, Chemical Engineering Department, Universiti Teknologi PETRONAS, Seri Iskandar, Perak 32610, Malaysia.
³ Department of Chemical and Environmental Engineering, University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor, Malaysia.
⁴ Chemical Engineering Department, Faculty of Engineering, Computing and Science, Swinburne University of Technology, Jalan Simpang Tiga, 93350 Kuching, Sarawak, Malaysia. Electronic address: bshow@swinburne.edu.my.
⁵ Department of Chemical Engineering, Faculty of Engineering and Science, Curtin University Malaysia, CDT 250, 98009 Miri, Sarawak Malaysia.

PMID: 31445240
DOI: 10.1016/j.biortech.2019.121971

Abstract

The aim of this study is to identify the optimum thermal conversion of Chlorella vulgaris with neuro-evolutionary approach. A Progressive Depth Swarm-Evolution (PDSE) neuro-evolutionary approach is proposed to model the Thermogravimetric analysis (TGA) data of catalytic thermal degradation of Chlorella vulgaris. Results showed that the proposed method can generate predictions which are more accurate compared to other conventional approaches (>90% lower in Root Mean Square Error (RMSE) and Mean Bias Error (MBE)). In addition, Simulated Annealing is proposed to determine the optimal operating conditions for microalgae conversion from multiple trained ANN. The predicted optimum conditions were reaction temperature of 900.0 °C, heating rate of 5.0 °C/min with the presence of HZSM-5 zeolite catalyst to obtain 88.3% of Chlorella vulgaris conversion.

Keywords: Artificial neuron network; Microalgae; Particle swarm optimization; Simulated Annealing; Thermogravimetric analysis.

MeSH terms

Catalysis
Chlorella vulgaris*
Microalgae*
Neural Networks, Computer
Temperature