Wheelchairs, as travel tools for the elderly and disabled, have enormous social demand. With the rapid development of technology, the wheelchair functions are unceasingly enriched and the related technologies are inproved, so high-performance and multi-function wheelchair robot has become an inevitable trend in the future. The wheelchairs, with flexible and comfortable mobile mechanism, and concise and intelligent control features, aslo become one of the current research hotspots.At present, when the commonly used wheelchair walks on the slope, the seat tilt will bring discomfort and especially tip-over phenomenon may happen in the large dip slope. Only with the assistance can their safety be ensured. In addition, when the wheelchair users take basic activitits in interior corridor or access in and out of a lift, and participate in social activities like visiting the exhibition and shopping in pedestrian streets, difficult turning problems take place frequently because of the narrow and crowed of the place. The main reasons for the problems is that the existing wheelchair usually adopts the structural form of two front support wheels and two back drive wheels, which will inevitably lead to large turning radius and is difficult to move flexibly in confined space, let alone omni-directional motion. In view of this, an omni-directional wheelchair robot with center of gravity self-adjusting mechanism, which not only can moves freely outside, but also can be used in narrow or crowded space, is presented. The main work of this paper includes:1) The omni-directional wheelchair robot system with center of gravity self-adjusting mechanism is planned. The construction method and technology implementation of the mechanism and control system is introduced with details. Based on ensuring comfortable seating, the functions of omni-directional motion and the center of gravity self-adjusting in real time are realized.2) The motion of omni-directional mobile platform is studied. Kinematic and dynamics model of the robot is established based on the analysis of a single differential unit. In addition, according to the dynamic model, trajectory tracking control strategy is analyzed. Trajectory tracking controller with stability and robustness is designed using adaptive sliding mode control with Lyapunov stability in consideration of interference.3) The center of gravity self-adjusting system is presented. The stability condition and dynamic model on the slope are analyzed. Based on accelerometer and gyroscope measurement principle, Kalman filtering method for multi-sensor information fusion is used to achieve efficient and accurate measurements of wheelchair pitch angle. Then, the center of gravity self-adjusting system is designed. The seat pose is adjusted by the actuator in real time to guarantee the stability and coordination of the wheelchair robot during the movement.4) For the unstructured and complex environment with pedestrians, autonomous navigation and avoidance system with multi-sensor detects is built. Based on the identification and analysis of external environment with the perceptual system consisting of laser scanners, camera and pose reference system, the method of creating a two-dimensional map from three-dimensional environment was proposed. Then, the adaptive Monte Carlo localization and dynamic pedestrian avoidance upon the map is realized.5) Finally, experiments on the omni-directional wheelchair robot with center of gravity self-adjusting mechanism are carried out, including flat walking experiment on the flat ground and on the slope, manned walking experiment and obstacle avoidance experiment under unstructured environments. These experiments have systematically verified the robot performance of mobility, center of gravity adjustment, climbing, manned walking, navigation and obstacle avoidance. Moreover, rationality and effectiveness of the theoretical analysis and technical method in the paper are validated. |