Research On Obstacle-avoidance System Of Performance Arm Of Intelligent-video Bridge-detection Vehicle

Bridge plays a very important role in joining different places and carrying the transportation in human history. While, with the increasing development of society and economy, the capacity of transportation and the load of bridge go up greatly, and the accidences of bridge collapse occur more than ever, which make the detection, maintenance and security of bridge become more and more attentive. Therefore, the guarantee of bridge's health has already been a safeguard of life and wealth of people and the study of detective technology has become an important domain.Nowadays, the detection of bridge's surface are very popular and easy to perform. On the other hand, due to the complexity of bridge's bottom, only one method which is manual periodical detection in the way of entrance to the bottom is popular. The method has a lot of shortcomings such as low efficiency and narrow adaptation, which makes it a good way to develop Intelligent-video Bridge-detection Vehicle (TVBDV) to realize autonomous detection which will overcome the fault of manual method and lead to progress in technology of bridge detection.In this paper, the emphasis is played on the study of the obstacle-avoidance system (OAs) of Performance Arm (PA) of IVBDV which controls and carries detective device such as camera along the bottom of bridge at a constant distance in a way of autonomous motion to avoid structural obstacles and cover almost everywhere. By reference to foreign and domestic fruit in OA method and system, an OAs of PA is put forward to concentrate on the specialty of the bottom of bridge and the autonomous detective method combining the autonomous OAs with nonholonomic OAs; based on inductive classification of detection system, a detective subsystem is designed to meet with the need to work in the open air using ultrasonic sensors as "external vision", infrared sensors as "external touch" and encoded sensors as "internal feeling", and also a Minimum Assessment (MA) method is used to infuse datum of multiple sensors to guarantee the detection effectively and real-time; a groping profile modeling control method and an electron-hydraulic neural net based control method are presented to realize OA control; at last, a simulated experiment system is set up to validate the research in above with the conclusion proving the OAs of PA effective and useful in the autonomous-detection field of bottom of bridge.