Fuzzy-Logic-Based Programming

The following sections are included:

• Table of Contents

• PREFACE

• ACKNOWLEDGMENTS

Most of the applications that will be discussed have the following general functional blocks:

INPUT → DATA ANALYSIS → OUTPUT

The INPUT functional block means to get input data at a given time. Input can be single input or multiple input. If it is a single input, a single value is obtained. If it is a multiple input, more than one value is accessed. Input data may be obtained automatically by taking some measurements of a physical system such as a machine, or by asking the user to manually enter data into a form in a database system. To measure a physical quantity, e.g., speed of a car, we use a sensor to convert it into an electrical signal which is then amplified by signal conditioning. The conditioned signal is finally digitized by a data acquisition board for a computer to process…

NICEL [Nicesoft 1994] is designed according to the principles of modularization, parameterization, packaging, reuse and a declarative (rather than procedural) approach in Software Engineering and Artificial Intelligence. Since NICEL allows the use of fuzzy logic and fuzzy rules in an intuitive manner, software coding in NICEL is far easier and more straightforward than coding in a conventional language.

In application areas such as laboratory automation, process monitoring and control, electronics testing, manufacturing, automotive and aerospace engineering, medical research, and educational instruction and research, the first thing that is necessary is hardware and software that measures and acquires physical phenomena data. Because of the wide spread of personal computers (PCs), we can use PCs as platforms for data acquisition. Figure 3-1 shows a general PC-based data acquisition system. All the components in this system can be purchased from various vendors. Therefore, it is very easy to set up a PC-based data acquisition system.

Today, the majority of fuzzy logic applications are in control. If you drive a new Saturn automobile with an automatic transmission, you are using fuzzy logic [Legg 1993]. Many Japanese cars use fuzzy logic in their anti-lock braking systems, cruise control, engine and emission systems, and active suspension systems. In Appendix F, some specific fuzzy logic applications in control are listed…

Pattern recognition is a field of classifying objects into classes or categories. For example, voice recognition classifies voices into words, and handwriting recognition classifies handwritings into letters, digits, or words. Each object (pattern) is usually represented by a set of feature (attribute) values…

An image is a 2-dimensional representation of objects captured from input devices such as video cameras, VCRs, document scanners, fax machines, film scanners, CCD video microscopes, X-ray machines, magnetic resonance scanners, infrared ray sensors, satellite remote sensors, etc. Mathematically, an image can be thought of as a matrix whose elements are called pixels. Each pixel can have a value that represents a gray level or a color at the pixel position…

There are formated and unformated data in databases. For example, tables in a relational database are formated data, where a relation is represented by a table whose columns correspond to attributes of the relation, and whose rows correspond to records. Examples of unformated data are text, waveforms, and images…

The purpose of data mining is to discover patterns in data. The necessary requiement of data mining is that you have data. Depending upon the application, the size of data can be small or large. For example, if the goal is to find rules for truck docking or car parallel parking, an expert driver may perform this task and data can be collected by sampling the sensor measurements of input and output variables at a fixed time interval. From the collected data, the subsequent algorithms can be applied to generate fuzzy rules. Usually, a data set collected for a control system is small…

Since the introduction of the NICEL language in [Nice soft 1994], it has been used to represent algorithms for many real-world applications. In this chapter, it will be explained how the Sugeno-Takagi-Kang model can be represented in NICEL. Furthermore, extensions of NICEL will also be discussed…

The following sections are included:

• Appendix A. Tranducers and Sensors Vendors

• Appendix B. Data Acquisition Boards Vendors

• Appendix C. Functions in LabWindows/CVI

• Appendix D. Functions in UniversalLibrary

• Appendix E. Functions in UniversalTools

• Apppendix F Some Applications In Control

• Appendix G Some Applications In Pattern Recognition

• Appendix H. LEADTOOLS DLL Functions

• Appendix I. GLIDE 2.1 Library Functions

• INDEX