1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Title:	Development of optimization model and artificial neural network technique for production bio-methanol via pyrolysis
Authors:	Nor Hazelah Kasmuri (P71230)
Supervisor:	Siti Kartom Kamarudin, Prof. Ir. Dr.
Keywords:	Biomass energy Neural networks (Computer science) Universiti Kebangsaan Malaysia -- Dissertations Dissertations, Academic -- Malaysia
Issue Date:	9-Dec-2018
Description:	Due to the depletion of petroleum reserves and the environmental impacts from fossil fuels, it is important to seek for an alternative source for energy. Biomass is one of the renewable resources of energy that offers solution in producing liquid fuel like bio-methanol. The potential biomasses that are able to produce bio-methanol are sugarcane bagasse, roots (wood), sawdust, trunk (wood), bark (wood), rice bran, palm oil residues and dry leaves. The objectives of this research are to characterize the physical and chemical activities of the raw materials, that includes the ultimate and proximate analysis like moisture content, volatile matter, ash content, fixed carbon and percentage of carbon, hydrogen, nitrogen, sulfur and oxygen elements in the sample. Secondly, this study develops model for material and energy balances as well as economic model for production of bio-methanol based on Runge Kutta ODE45 method and optimized using Matlab® (ver. R2011a). The models are verified using experimental data. Fast pyrolysis of biomass consists of pyrolysis reactor (R1) and condenser (C1) operated batchwise used for the production of bio-methanol. The reactor is operated at a temperature range of 400 to 500°C. Design of experiments and optimization are conducted based on face centered-central composite design (FCCCD) and response surface method (RSM) to determine the significant parameters affecting the process. In findings, bio-methanol was highly derived from sugarcane bagasse feedstock with 3.11 wt. % in yield. The quadratic model fit to the study where coefficient determination, R2 and adjusted R2 were evaluated at 0.88 and 0.77, respectively. Results show that the optimum operating conditions were found at 500°C, 120 min and 2 l/min. Whereas the yield of bio-methanol predicted and actual data was determined at 3.075 wt. % and 2.876 wt. %, respectively. Finally, based on the significant parameters obtained, this study developed the control system. The dynamic control system was developed and simulated. The sustainability concern of bio-methanol involved the integration of ANN-Simulink control strategy applied with an artificial neural network (ANN) technique in NN control toolbox. Model reference control in dynamic study of nonlinear bio-methanol production regarding the process design and control parameters was employed. The significant parameters that influence the process are namely the reaction time, reaction temperature and nitrogen flow operated in pyrolysis batch reactor. The objective is emphasised for maintaining high yield of bio-methanol production in pyrolysis with set point constant values at optimum conditions from experimental studies. Thus, the development of model reference control with neural controller has attained a better control in manipulating of reaction temperature for pyrolysis batch reactor. The predicted data from ANN fitting model was determined with validation, R2 achieved at 0.98 for nonlinear quadratic model of bio-methanol equations. The bio-methanol production yield attained 3.09 wt. % in prediction value. MSE error was indicated at small difference of 0.2617 in validation result. Therefore, the non-linearity of regulating input temperature to linearity of measured output of bio-methanol yield was achieved in this study.,Ph.D.
Pages:	202
Call Number:	TP339.N635 2018 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.