EXERGY PARAMETRIC ANALYSIS AND PREDICTION OF TURMERIC RHIZOME SLICES DRYING USING NEURO-FUZZY, NEURAL-NETWORK AND REGRESSION TECHNIQUES

E. O. Oke; B. I. Okolo; O. Adeyi; J. A. Otolorin; J. A. Adeyi; C. J. Ude; G. W. Dzarma; K. N. Akataobi; F. Nwokocha

E. O. Oke Chemical Engineering Department, Michael Okpara University of Agriculture, Umudike, Abia State State, Nigeria
B. I. Okolo Chemical Engineering Department, Michael Okpara University of Agriculture, Umudike, Abia State State, Nigeria
O. Adeyi Chemical Engineering Department, Michael Okpara University of Agriculture, Umudike, Abia State State, Nigeria
J. A. Otolorin Chemical Engineering Department, Michael Okpara University of Agriculture, Umudike, Abia State State, Nigeria
J. A. Adeyi Mechanical Engineering Department, Ladoke Akintola University of Technology, Ogbomoso, Oyo State, Nigeria
C. J. Ude Chemical Engineering Department, Michael Okpara University of Agriculture, Umudike, Abia State State, Nigeria
G. W. Dzarma Department of Chemical Engineering Michael Okpara University, Umudike, Nigeria
K. N. Akataobi Department of Chemical Engineering Michael Okpara University, Umudike, Nigeria
F. Nwokocha Department of Chemical Engineering Michael Okpara University, Umudike, Nigeria

Keywords: exergy-sustainability, exergy efficiency, exhaustive-search and drying

Abstract

This study presents exergy parametric analysis and prediction of turmeric rhizome drying using first and second law of thermodynamics as well as soft-computing techniques. The drying experiments were conducted at inlet drying temperature: (40-650C), air velocity (1.5-3m/s), drying time: (30-240 minutes) and sample thickness: (2-5mm). The Neuro-Fuzzy Exhaustive Search (NFES) parametric analysis results revealed that drying time (RMSE=0.0031), temperature (RMSE=0.096), temperature (RMSE=0.046) and sample thickness (RMSE=0.748) are the most single relevant parameters for Exergy Loss (EL), Exergy Efficiency (EE), Exergetic Improvement Potential (EIP) and Sustainability Index (SI) respectively. Whereas temperature-time (RMSE=0.0031), temperature-velocity (RMSE=0.0945), temperature-time (RMSE=0.046) and time-thickness (RMSE=0.7534) are the most important two-input combinations for EL, EE, EIP and SI correspondingly. NFES also revealed that time-temperature-velocity (RMSE=0.004), temperature-velocity-thickness (RMSE=0.082), time-temperature-velocity (RMSE=0.0436) and time-temperature-thickness (RMSE=0.758) are the three-input significant combination for EL, EE, EIP and SI respectively. The ANN results show that two-input combination architectures gave the highest R2 with minimum RMSE for the exergy-sustainability indicators. Therefore, this study shows that NFES and ANN are reliable tools for the analyses of turmeric rhizome drying thermo-sustainability indicators.