Управление процессом безударного захвата непредсказуемо движущегося объекта мобильным манипуляционным роботом

Abstract

An algorithm is constructed for controlling a non-impact capture of unpredictably moving target by the robotic arm. Capture is performed in a finite time. The solution is obtained without the use of the information on the non-control forces, including the disturbing forces and forces of inertia. The object is achieved in four stages. First, the principal vector of the forces is obtained, which provides movement of the center of mass of the robot body in the mode of persecution on the basis of proportional navigation in the pursuit of the object. Second, the principal moment of the forces about the center of mass of the body is obtained, which provides bringing of the one of the principal central axes of inertia of the moving coordinate system associated with the robot’s body, in a position coinciding with the line of sight. Third, an additional driving force is determined, which provides the unstressed bringing of the attachment point of the first link of the arm with a robot body at a distance of “manipulator arm” from the target on the line of sight to provide capture. Fourth, the expression of forces and moments is constructed for the management of the translational and rotational motion of the links of the manipulator relative to each other, allowing bumpless capture of the pursued object. The self-adjusting method is proposed to automatically select the optimal values of the control. It is carried out by the “principle of feedback on the quasi-acceleration” at discrete points in time. This principle was first proposed by I.A. Mukhametzyanov in an article published in the Bulletin of Peoples’ Friendship University series “Mathematics. Information Sciences. Physics”, No 3, 2013.