الفهرس 
SummaryOnly 14 pages are availabe for public view
 |  | from | 144 |  |  |
| | | from | 144 | | |
Abstract
In the past decade there has been a growing interest for the development of olfactory machines and electronic nose systems. This surge has been mainly driven by a variety of real life applications. Indeed, the problem of classifying and further quantifying chemical substances on a real-time basis is very critical for a wide range of applications in the industrial and civil environments. The ability to monitor and precisely measure leakages of combustible and explosive gases is crucial in preventing the occurrence of accidental explosions. Accordingly, the development of gas sensors for the detection of single gases (such as CO, CH4, H2, SO2, NOx, O3 etc.) has seen an increasing interest within the research community. Environmental applications including air quality control and pollution monitoring are experiencing a steadily increasing attention as the need to protect the environment has grown over the last years. Quality control and agricultural applications (food, drinks agricultural products quality control, soil contamination, etc.) are also on the rise. New and exciting developments have also recently emerged in the area of medical applications (detection of infections, diseases and bacteria), mobile robot navigation as well as space applications (monitoring of air quality and gas detection in space shuttle). SO, the last decade has witnessed an increasing interest in modeling the biological olfactory systems and building bio-inspired electronic nose. Microelectronic gas sensors in particular, have received an increased attention in recent years, due to their numerous advantages including small size, high sensitivities in detecting very low concentrations, possibility of on-line operation and low-cost fabrication, making them attractive for consumer applications. Unfortunately, disadvantages of all gas sensors suffer from a number of shortcomings such as lack of selectivity, non linearities of the sensor’s response and long-term drift. In fact, an array of different gas sensors is used to generate a unique signature for each odor. A gas sensor array permits to improve the selectivity of the single gas sensor, and shows the ability to classify different odors and to quantify components concentrations. Pattern recognition algorithms combined with a gas sensor array have been traditionally used to address these issues. The aim of the pattern recognition techniques is to find a relationship between the sensors outputs and the odor class (or concentration). In order to achieve this, first some features have to be extracted from the sensors responses and then the functional relationship between the feature vectors and the class labels has to be derived by a learning procedure. Therefore, an electronic nose consists of an array of gas sensors, signal preprocessing, and a pattern recognition algorithm. So, the algorithmic part of an odor discrimination system consists of three steps: (i) Signal conditioning and feature extraction The role of the first step is to segment the pattern of interest from the background, remove noise, normalize the pattern, and any other operation that contributes in defining a compact representation of the pattern. (ii) Feature reduction Feature reduction should provide a small number of informative features in order to make the learning task simpler. (iii) Classification Classification tasks address the problem of identifying unknown sample as one from a set of recognizable gases. In summary, the aim of this work is to make a comparison between different pattern recognition techniques in order to find a relationship between the sensors outputs and the odor class (or concentration) by using a reliable and robust pattern recognition algorithms to address the learning problem in electronic noses for identifying new odors. Also it is required to ensure that a designed electronic nose has good performance through introducing a suggested controller, based on an intelligent artificial neural network (ANN). The present thesis consists mainly of five chapters which can be summarized as follows: In chapter one Chapter one is an introduction about the electronic nose, its construction, performance and its evaluation. In addition this chapter illustrates the main objective of the thesis and ends by a summary about the thesis contents. In chapter two A historical review of the development of E-Nose and its applications in different fields. Also this chapter gives a brief introduction about the chemical sensors used with electronic nose. In chapter three A general review of neural networks including its architecture, types of learning, and algorithms are also introduced. This chapter also discusses the suggested neural network algorithm for controlling E-Nose. In chapter four This chapter discusses the case study and the application of the suggested algorithm and the results obtained with a comparative analysis between neural networks. In chapter five This chapter contains a conclusion for the whole results and suggestions for future work.