Two approaches to improve robot capabilities

The paper presents two approaches to instil advanced capabilities to the industrial and mobile robot. The first one is related to the redundant industrial robot, motion of which is to be optimized under kinematic restrictions caused by the robot operation in a tight space. To solve the task the solution of inverse kinematics problem was converted into a closed-loop optimization problem, due to which the control algorithm provided high precise and robust solution w.r.t outer influences. The algorithm was verified by simulation and at present runs its implementation on ABB IRB 4600 robot. The second one is related to learning fuzzy-neural navigation of a mobile robot in unknown environment cluttered with obstacles. The learning process runs in two modes. First the parameters of membership functions (MF) are updated, and then the walls of MFs are adaptively deformed so as to remove transversal swings from the robot trajectory. The navigation algorithm was verified by simulation and validated by the real mobile robot.