LINEAR DISCRIMINANT ANALYSIS BASED SENSING CHARACTERISTICS STUDY OF ZnO THICK FILM GAS SENSOR

Driving under the influence (DUI) is a synonymous term that represents the criminal offense of operating a motor vehicle while being under the influence of alcohol. Semiconductor oxides such as SnO2, TiO2, and ZnO have been more successfully employed as sensing materials compare to organic semiconductors for the detection of ethanol gas concentration. This work investigates the fabrication and characterization of the thick-film gas sensor based on Zinc-oxide on an Alumina substrate. The sensor exhibits excellent ethanol sensing characteristics at temperatures between 175°C to 300°C and the characteristics of the sensor for different concentration of the ethanol gas has been successfully studied using Linear Discriminant Analysis (LDA). This system can be used to detect alcoholic driving for automotive safety applications. The system is designed to detect the exact concentration of ethanol gas using virtual instrumentation software and data acquisition hardware. This work presents the data analysis results of 230 samples of each concentration levels of ethanol gas using Linear Discriminant Analysis (LDA) technique. The overall system can be tested with the engine ignition system for automotive safety and warning applications in the prevention of a crash during drunken driving. Keywords-Linear Discriminant Analysis (LDA); Engine ignition System; Zinc Oxide Thick film sensor; Driving under the influence (DUI); Virtual Instrumentation. INTRODUCTION Driving under the influence (DUI) (drinking and driving, and drink-driving) is the act of operating any motorized machinery after or during the consumption of alcohol or other drugs. DUI is synonymous term that represents the criminal offense of operating a motor vehicle while being under the influence of alcohol or drugs or a combination of both. It is a criminal offense in most countries. This system will detect and display ethanol concentration in the LabView front panel. The sensor is a thick film based zinc oxide gas sensor. The sensor is having an inner heating coil which is heated up to a temperature range of 75°C to 300°C. The use of resistive, adsorption-based sensors has been increasing over the past few years for the purposes such as detection of smoke, oxidizing or reducing gases (O2, CO, Ethanol, CH4, CO2, etc) and humidity [1]. A large number of materials have been used for gas sensing, including ceramics that often consist of a combination of metal oxides. ZnO is one of the semiconductor materials that are sensitive to many gases of interest like hydrocarbons, hydrogen, volatile organic compounds etc. They exhibit high sensitivity, satisfactory stability and rapid response to even small concentrations of gases (ppm level). In this paper we will concern ourselves with the sensitivity of ZnO thick film resistor for sensing ethanol gas at 100ppm. of gas concentration at various temperatures from 75°C to 300°C in order to find the temperature for maximum sensitivity [2]. Also the sensitivity for different concentrations of ethanol (from 5ppm. to 95ppm.) was tested. The sensor characteristics can be studied using various data clustering techniques like Linear Discriminant analysis (LDA). In this paper we have done the analysis using Linear Discriminant analysis (LDA). This sensor can easily be implemented as an engine ignition system during drunken driving [Figure 1]. Fig. 1Zinc Oxide Gas Sensor used for detection and warning of DUI and engine ignition system. Linear Discriminant Analysis Based Sensing Characteristics Study of ZnO Thick Film Gas Sensor 32 Advances in Computational Research ISSN: 0975–3273 & E-ISSN: 0975–9085, Volume 3, Issue 1, 2011 A. Linear Discriminant Analysis (LDA): Linear Discriminant Analysis (LDA) is a commonly used technique for data classification and dimensionality reduction. Linear Discriminant Analysis easily handles the case where the within-class frequencies are unequal and their performance has been examined on randomly generated test data. This method maximizes the ratio of between-class variance to the within-class variance in any particular data set thereby guaranteeing maximal separability. The use of Linear Discriminant Analysis for data classification is applied to classification problem in speech recognition. Data sets can be transformed and test vectors can be classified in the transformed space by two different approaches. In this type of LDA, each class is considered as a separate class against all other classes. The class category or the group (“equal” and “not equal”) is what we are looking for (it is also called dependent variable). Each measurement on the product is called features that describe the object (it is also called independent variable). Thus, in discriminant analysis, the dependent variable (Y) is the group and the independent variables (X) are the object features that might describe the group. The dependent variable is always nominal category variable while the independent variables can be of any measurement scale i.e. nominal, ordinal, interval or ratio. If we can assume that the groups are linearly separable, we can use linear discriminant model (LDA). Linearly separable suggests that the groups can be separated by a linear combination of features that describe the objects. If only two features, the separators between objects group will become lines. If the features are three, the separator is a plane and the number of features (i.e. independent variables) is more than three, the separators become a hyper-plane. The classification rule is to assign an object to the group with highest conditional probability. This is called Bayes Rule. Fortunately, there is a relationship between the two conditional probabilities that well known as Bayes Theorem.