No abstract

We propose a novel shared control strategy for mobile robots in a human-robot interaction manner based on surface eletromyography (sEMG) signals. For security reasons, an obstacle avoidance scheme is introduced to the shared control system as collision avoidance guidance. The motion of the mobile robot is a resultant of compliant motion control and obstacle avoidance. In the mode of compliant motion, the sEMG signals obtained from the operator's forearms are transformed into human commands to control the moving direction and linear velocity of the mobile robot, respectively. When the mobile robot is blocked by obstacles, the motion mode is converted into obstacle avoidance. Aimed at the obstacle avoidance problem without a specific target, we develop a no-target Bug (NT-Bug) algorithm to guide the mobile robot to avoid obstacles and return to the command line. Besides, the command moving direction given by the operator is taken into consideration in the obstacle avoidance process to plan a smoother and safer path for the mobile robot. A model predictive controller is exploited to minimize the tracking errors. Experiments have been implemented to demonstrate the effectiveness of the proposed shared control strategy and the NT-Bug algorithm. INDEX TERMS Shared control system, surface electromyography, no-target Bug algorithm, model predictive control.

No abstract

This paper proposes a hybrid path planning method based on artificial potential field method (APF) for mobile robot, which combines wall following method (WFM) and obstacles connecting method (OCM) for dealing with local minimum. The environment information is took into consideration to decide the escape direction of WFM. To ensure the success of escaping from local minimum, more reliable switching conditions are designed. OCM is applied to reduce the difficulty of path planning for complex workspace with concave obstacles. Simulation studies have been carried out to verify the validity of the proposed method.

No abstract