Contents lists available at ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel Full Length Article Prediction of engine performance and emissions with Manihot glaziovii bioethanol - Gasoline blended using extreme learning machine A.H. Sebayang a,b , H.H. Masjuki a , Hwai Chyuan Ong a, ⁎ , S. Dharma a,b , A.S. Silitonga b,c,d , F. Kusumo a , Jassinnee Milano a a Department of Mechanical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia b Department of Mechanical Engineering, Politeknik Negeri Medan, 20155 Medan, Indonesia c Department of Mechanical Engineering, Syiah Kuala University, 23111 Banda Aceh, Indonesia d Department of Mechanical Engineering, Faculty of Engineering, Universiti Tenaga Nasional, 43000 Kajang, Selangor, Malaysia ARTICLE INFO Keywords: Engine performance Exhaust emission Manihot glaziovii bioethanol Extreme learning machine Alternative fuel ABSTRACT Bioethanol can potentially replace gasoline because of its lower exhaust emissions. The purpose of this study was to investigate the engine performance and exhaust emissions of Manihot glaziovii bioethanol–gasoline blends at different blend ratios (5%, 10%, 15%, and 20%). Tests were performed on a single-cylinder, four-stroke spark- ignition engine with engine speed was varied from 1600 to 3400 rpm, and the properties of the Manihot glaziovii bioethanol–gasoline blends were measured and analysed. The vapour pressure increased for fuel blends with low concentrations of bioethanol due to the oxygen within the bioethanol molecules and the contribution of the flame speed which can enhance the combustion and improved the engine efficiency. In addition, the engine torque, brake power, and brake-specific fuel consumption (BSFC) were measured, as well as the carbon mon- oxide (CO), hydrocarbon (HC), and nitrogen oxide emissions. For a fuel blend containing 20% bioethanol at an engine speed of 3200 rpm, the BSFC decreased, with maximum values of 270.7 g/kWh. The CO and HC emis- sions were lower for the Manihot glaziovii bioethanol–gasoline blends. In addition, an extreme learning machine (ELM) model was developed for application in the automotive and industrial sectors. This tool reduces the cost, time, and effort associated with experimental data. The blend ratio of the bioethanol–gasoline blends and the engine speed were used as the input data of the model, and the engine performance and exhaust emissions parameters were used as the output data. The coefficient of determination (R 2 ) was within a range of 0.980–1.000, and the mean absolute percentage error was within a range of 0.411%-2.782% for all the parameters. The results indicate that the ELM model is capable of predicting the engine performance and exhaust emissions of bioethanol–gasoline fuel blends. 1. Introduction Bioethanol is a promising source of energy for replacing gasoline in the future [1,2]. Gasohol is an alternative fuel produced by blending gasoline with bioethanol. Bioethanol has the following advantages over gasoline: a higher octane number, a broader flammability limit, a higher flame speed, and a higher heat of vapourisation [3,4]. The fa- vourable properties of bioethanol result in a higher compression ratio, a shorter burn time, and a leaner burning engine. Generally, bioethanol leads to complete combustion in spark-ignition engines, in contrast to gasoline. New alternative fuels have been developed for use in spark-ignition engines with a higher thermal efficiency to satisfy the stringent emis- sion regulations in recent years [5,6]. Considerable research has been conducted to investigate the effects of bioethanol on the performance and exhaust emissions of spark-ignition engines [7,8]. Najafi et al. [9] investigated the performance of a four-stroke spark-ignition engine fuelled with gasoline–ethanol blends, finding that the combustion ef- ficiency was improved by 35% for a fuel blend containing 20% ethanol. However, the results showed that the fuel consumption in case of the gasoline–ethanol blend was lower than that when using gasoline as the fuel. Koç et al. [10] investigated the performance of a single-cylinder spark-ignition engine fuelled with gasoline–ethanol fuel blends con- taining a high percentage of ethanol (50% and 85%), reporting that the brake-specific fuel consumption (BSFC) decreased by 20.3% and 45.6% for blends containing 50% and 85% ethanol, respectively. Ghazikhani et al. [11] analysed the engine performance and emissions for gasoli- ne–ethanol fuel blends. Their results showed that the carbon monoxide http://dx.doi.org/10.1016/j.fuel.2017.08.102 Received 31 March 2017; Received in revised form 24 August 2017; Accepted 29 August 2017 ⁎ Corresponding author. E-mail addresses: ong1983@yahoo.com, onghc@um.edu.my (H.C. Ong). Fuel xxx (xxxx) xxx–xxx 0016-2361/ © 2017 Elsevier Ltd. All rights reserved. Please cite this article as: Sebayang, A.H., Fuel (2017), http://dx.doi.org/10.1016/j.fuel.2017.08.102