Task Oriented Motion Planning Of An Intelligent Wheelchair Robot For Disabled And Elderly

With the aging of Chinese, assisting handicapped or elderly people has become an important issue to be resolved. The intelligent wheelchair service robot for the disabled and elderly is a robot that an intelligent wheelchair equipped with a manipulator. The robot combines the mobility of the wheelchair and manipulability of the robot arm that assist the activities of daily living for the disabled and elderly and insure their basic viability in the case of out of care. Now the control of existing commercial robot arm for wheelchair is manual control and the arm and the wheelchair are the two independent system, this greatly increases the cognitive load for the user. So it is necessary to put the wheelchair and the robot arm to a whole system to study the intelligent control. According to the need of the control of the whole system, the paper put the 2 degree of freedom wheelchair and the 6 degree of freedom robot arm as the redundant robot. The motion planning of the robot is studied in this paper Ten activities of daily livings are planned reasonably. According to the different characteristics of the tasks, the trajectory planning method of the subtasks in the joint space is studied. The concrete research content is as follows:Firstly, the JACO robot arm which is produced by Kinova company is choosed. The kinematics of the arm is analyzed According to the three wrist joints are non-orthogonal, the algebraic elimination method based on polynomial method is used to obtain the solution for inverse kinematics. The workplace of the arm and the manipulability in the workplace are analyzed.Secondly, 8 DOF redundant robot which is combined by the 2 DOF wheelchair and the 6 DOF robot arm is established. The kinematics of the redundant robot is analyzed. The improved least-norm solution method which make joint limit avoidance as objective function is used to obtain the optimized solution for inverse kinematics.Then the motion planning of the arm is studied based on the a ctivities of daily livings. Evaluation index of the task completion is established. 10 activities of daily livings are defined and the basic motion database is established. The complicated activities of daily livings could be broken into subtasks by the basic motion. Then the trajectory of the subtasks is planned based on the minimum time-jerk algorithms in the joint space. The algorithms is validated by simulation. Finally, complete the motion planning of the 10 activities of daily and the motion planning method is validated by the experiment of execute the drinking task.