Study On Intelligent Anti-collision System For Expressways

Intelligent Transportation System (ITS) is a kind of systems be combined with the advanced computer technology, information technology, data-communication technique, auto-control art, artificial intelligence and electronic technology, etc, It has been applied in traffic-transportation management system, helps to establish a wide-extent, whole-direction, accurate, punctual, efficient traffic-transportation management system. Over the past two decades, Intelligent Transportation System has experienced rapid development in the United States, European countries and Japan, achieving significant economic and social benefits. Nevertheless,development of ITS is still at a preliminary stage in China.In recent years, China has invested a large amount of money for the construction of transportation infrastructure. There has been accelerated development of the transportation infrastructure. After nearly 20 years of efforts, China's expressway construction has started from scratch and has yielded fruitful results. Despite the rapid development in recent years, the increasingly frequent traffic accidents are still causing a large number of casualties and huge economic losses. How to reduce or avoid traffic accidents on expressways as well as make rapid judgment and quick response after traffic accidents happened have been hot topics in the field of intelligent transportation both at home and abroad.This paper aims at the research of rear-end collision alarm system based on ITS. The purpose is to provide a simple and inexpensive anti-collision system. The system consists of two parts: the anti-collision system based on digital scale, the data transmission system based on GPRS. We used this system in JinJi expressway for one mile and achieved good experimental results. But we still need to improve the local optimization.This thesis discusses the design principle of the above two systems, analyses the theory problems involved in our design. Hardware and software design for the establishment are described in detail. The lane recognition algorithm of the frequency domain based on the discrete cosine transform is described in the last part.