| In recent years, the control of Wheeled Mobile Robot (WMR) trafficability with drive failure has become the popular field in robotic research. This thesis is supported by Harbin Special Funds for Technological Innovation Research Projects named "the ability of fault-tolerant toward WMR in the state of motion", and it proposes a control method and reconfiguration to keep the trafficability of WMR when the drive failure situation happens.Firstly, on the uneven terrain, considering the situation of drive failure, the remaining driving wheels may still have enough power to drive the WMR. So, in order to complete scheduled tasks for the WMR with some wheels driving failure,it's necessary to have a quasi-force analysis to know the impact of failure. According to different cases of failure,it proposes a method of configuration transformation to utilize the remaining driving power effectively. Based on the Quadratic Programming algorithm, the equation from quasi-force analysis is solved to judge the trafficability on the rough ground and simulated in MALAB.Secondly, nonlinear control of mobile robot is studied. Considering WMR's inevitable slip on the uneven ground, which makes the accurate mathematical model cannot get, a nonlinear control method with strong robust is needed. Base on the idea of Backstepping,conduct variable structure control to trajectory tracking of WMR.Thirdly, binocular vision is adopted as visual sensor of WMR and visual feedback is designed to position machine of WMR. The information collection, analysis and control of binocular visual are programmed in Lab VIEW.Finally, conducting the path following experiment to different cases of failure on the flat and uneven ground respectively, the information of wheels current, IMU and tracking error are collected in the experiment and consequently the results are analyzed to verify the effectiveness of proposed method to driving failure. |
PDF Full Text Request