Abstract
In this paper, a review is given of some of the modern methods in the detection of knock in internal-combustion engines and
some comparisons are made between these methods and the effectiveness of each one of them is indicated through a statement of
the advantages and disadvantages of each method. In this way it will be possible to clarify how to deal with the original signal
and the associated signal noise through some of the modern algorithms in the field of soft computing such as an Artificial Neural
Network (ANN), Genetic Algorithms (GA), Wavelet Transform (WT), Fuzzy logic, Supported Vector Machine (SVM) and some
statistical methods.
© 2013 The Authors. Published by Elsevier B.V.
Selection and peer-review under responsibility of the Faculty of Information Science and Technology, Universiti Kebangsaan
Malaysia.
Keywords: Artificial Neural Network; Fuzzy logic; Genetic Algorithms; Supported Vector Machine; Wavelet Transform.
1. Introduction
Knocking is a process that presents a challenge for many engineers and researchers to achieve the characteristics
of quality and to meet customer satisfaction through the achievement of efficiency in the engine. A design that
avoids the occurrence of knocking is not an option, but this involves the use of high cost excellent quality fuel, a low
compression ratio, and high-efficiency control strategies to avoid knocking. It is important to manipulate the
compression ratio and spark advance to accommodate knocking, as the knocking process leads to
* Corresponding author. Tel.: +60174471347
E-mail address: azherwitwit@yahoo.com