Bridge Inspection Climbing Robot And Adaptive Control Technology

Bridge accidents happened frequently both domestic and overseas in recent years, bridge safety is getting increasingly attention of the whole society. With the physical and chemical erosion, a variety of diseases have occurred on bridges. Traditional bridge inspection methods are mainly by naked eyes or bridge-inspection vehicle. These ways always require much time and effort, and are dangerous to inspectors. So a novel and efficient bridge inspection methodology is increasingly required.This paper designed and developed a bridge inspection system of climbing robot, which is applied for bridge environment. The robot is worked as a fast-moving platform, on which a series of sensors are integrated and a wireless visual inspection device is equipped. Inspector operates robot through a remote controller and it can return detected images to ground workstation. Then the bridge crack character is recognised by image-processing procedure online and store to database for danger level assessing. At last, engineer do corresponding measures by the appraised results.First, this paper analyzes and summarizes the current research of bridge inspection methods and climbing-robot application development. The overall system framework is designed. Robot kinematics modeling and adhesion secure conditions are analyzed following. Then an adaptive adhesion chamber is optimized. Based on these, the system hardware platform is built and related software is designed to realize bridge inspection tasks.Secondly, adhesion adaptive control methodology is applied to solve the problem that suction chamber is easily leaked according to diverse roughness wall. It is also focused contents of this paper. PID control and fuzzy control theories are discussed, then the robot adhesion system is modeled and simulated in MATLAB environment. The experimental results show the fuzzy PID controller has better adaptability and performance.Finally, various roughness wall experiments and bridge field experiments are performed by using the developed prototype robot. The results verify that the robot has strong adaptability to kinds of walls, and the entire system is reliable and efficiency. As an important basis and guidance, the online bridge crack recognition information is applied for bridge danger level assessment.