The turning motion of a tracked pipeline-robot implemented by skid steering is a nonholonomic dynamic problem with intrinsic nonlinearity, to which the classical control method is inappropriate and cannot be applied. ...The turning motion of a tracked pipeline-robot implemented by skid steering is a nonholonomic dynamic problem with intrinsic nonlinearity, to which the classical control method is inappropriate and cannot be applied. This paper presents a novel path tracking control method based on hierarchical fuzzy structure. The controller consists of three sub-level low dimensional fuzzy control systems： fuzzy supervisory control, fuzzy steering and fuzzy velocity control. As a result, establishing the bases of rules for the fuzzy control becomes feasible and simplified, and the related controller can be adapted to complicated ground and environment. Using this method, the number of fuzzy controt rules is greatly decreased so that the curse of dimensionality causing the multivariable problem does not occur. Simulation and experimental results validate the effectiveness of the proposed method with satisfied performance on path tracking. Autonomous navigation of the caterpillar-inspired pipeline-robot is also implemented based on the sensor feedbacks.展开更多
基金Supported by the National Science Foundation of China （ No. 60935001 ） and New Century Excellent Talents in University of Chinese Ministry of Education （NCET-06-0398）.
文摘The turning motion of a tracked pipeline-robot implemented by skid steering is a nonholonomic dynamic problem with intrinsic nonlinearity, to which the classical control method is inappropriate and cannot be applied. This paper presents a novel path tracking control method based on hierarchical fuzzy structure. The controller consists of three sub-level low dimensional fuzzy control systems： fuzzy supervisory control, fuzzy steering and fuzzy velocity control. As a result, establishing the bases of rules for the fuzzy control becomes feasible and simplified, and the related controller can be adapted to complicated ground and environment. Using this method, the number of fuzzy controt rules is greatly decreased so that the curse of dimensionality causing the multivariable problem does not occur. Simulation and experimental results validate the effectiveness of the proposed method with satisfied performance on path tracking. Autonomous navigation of the caterpillar-inspired pipeline-robot is also implemented based on the sensor feedbacks.
