An Embedded Radial Basis Network Structure Application onto Light Source Conjunction System

• Main Authors: Jang Wen Ben, 江文彬
• Other Authors: 胡永柟
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/65415002854003928687

碩士 === 大葉大學 === 電機工程學系碩士班 === 92 === ABSTRACT In nature there are many insects with positive photo taxis , i.e., they tend to seek the point light source. The process of their response to the external world is something like this: detect the variation in light intensity -- learn how to control from past experience — establish response pattern. In our study we borrowed this model and propose an embedded radial basis network structure that simulates the positive photo taxis of insects to design for a light source conjunction with an embedded system using a single chip microprocessor. Our design will light source. Perform control functions within the system to correct errors and learn to make the appropriate response. Our design will then generate response in the shortest possible time and consequently make the most efficient control. The content of this paper are organized into three parts: (1) Use the theory of artificial neural radial basis network structure to design an embedded system based on a single chip microprocessor. This light source tracking system has memories and the ability to learn. (2) Use light-sensitive resistors to detect the intensity of light. We use the photosensitive control chip in the system is the ATEMGA 8L. It performs control functions and sends out the response signal after making appropriate judgment based on past experiences. (3) We will use the C language conforming to the ANSI standard to write ICCAVR software. This software is compatible with the AVR single chip microprocessor, so it is an ideal driver for the microprocessor that automatically tracks the light source. The design outlined in this research paper combines theory and our experience in practical design. We are confident it can contribute to the applications in the field of automatic control. Particularly, we hope it will be used in the robotic design. Key Words: artificial neural radial basis network structure, positive photo taxis, AVR single chip microprocessor