Adaptive neuro-fuzzy system (ANFIS) based appraisal of accumulated heat from hydrogen-fueled engine. (6th July 2015)

• Record Type: Journal Article
• Authors: Taghavifar, Hadi; Khalilarya, Shahram; Jafarmadar, Samad
• Is Part Of: International journal of hydrogen energy. Volume 40:Number 25(2015)
• Page Start: 8206
• Page End: 8218
• Publication Date: 2015-07-06
• Subjects: Hydrogen as fuel -- Periodicals; Hydrogène (Combustible) -- Périodiques; Hydrogen as fuel; Periodicals; 665.81
• DOI: 10.1016/j.ijhydene.2015.04.089
• Languages: English
• ISSNs: 0360-3199

Abstract

<abstract xml:lang="en" abstract-type="author" id="abs0010"> <title id="sectitle0010">Abstract</title> <sec> <p id="abspara0010">Hydrogen as an ecological fuel has called a great deal of attention and has been a dynamic field of studying interest for engine designers. The work presented herein is an investigation of the accumulated heat created in hydrogen fueled engine as affected by crank angle, equivalence ratio, penetration and H<sub>2</sub> mass fraction. The capability of computational fluid dynamics (CFD) approach was applied to simulate the combustion process while the potential of a universal approximator, ANFIS (adaptive neuro-fuzzy inference system), was utilized to model the problem. Two statistical performance criteria of root mean square error (<italic>RMSE</italic>) and coefficient of determination (<italic>R</italic><sup><italic>2</italic></sup>) were considered to further assess the modeling performance. The findings revealed that due to high diffusivity characteristics of hydrogen, the optimum level of equivalence ratio is affected by the addition of hydrogen. It was observed that hydrogen addition yields higher heat release since hydrogen and air can mix well at any standard scale. Higher engine speed than that of 3500 rpm resulted in a higher spray penetration at any specified crank-angle position. It was also concluded that an ANFIS structure with <italic>Trimf</italic> membership algorithm each at three membership functions yielded <italic>RMSE</italic> of 0.011527 and R<sup>2</sup> equal to 0.999.</p> </sec> </abstract>