Abstract
Human-Robot Collaboration (HRC), as an important role in intelligent manufacturing, is widely utilized in modern factories. Mobile manipulators offer greater flexibility due to the ability to simultaneously move their bases and arms. This provides them an advantage facing challenging tasks such as flexible production and intelligent manufacturing. The environment of modern factories is complex, with slopes and thresholds at the entrance. Stability plays a fundamental role in complex environments. Therefore, for the future deployment and application of mobile manipulator in HRC, a novel dynamic tip-over avoidance method is proposed for mobile manipulators based on the extended zero moment point (ZMP) algorithm. This method extends the traditional ZMP stability criterion into three dimensions, deriving a system-wide three-dimensional zero moment point. The integration of the extended ZMP with redundant features facilitates adaptive weight matrix adjustments, enabling the robot to adapt its motion based on environmental constraints, thus preventing instability such as tip-over. Simulation verification is carried out with CoppeliaSim. Experimental verification is performed using the MR2000+Franka mobile manipulator. The results confirm that the extended ZMP algorithm effectively prevents tipping in complex environments.