Nonlinear Position Control Of Robot Manipulators In Task Space

In most practical situations motions of robot manipulators with revolute joints arespecified in robot task space. Normally, it is required to formulate the task spacecontrol scheme directly. However, the uncertain measurements of robot end-effectormay lead to very complex control system.At first, a new class of nonlinear proportional plus derivative control withreal-time gravity compensation (NP-D) of robot manipulators in task space is proposedto give a faster response and higher position control precision over the commonly usedcontrol strategies. The proposed NP-D is formulated with a class of nonlinear saturatedfunction with the characteristics of ‘enlargement of small error and saturated in largeerror’.Second, in practice, velocity measurements are prone to be contaminated by noiseand degrade the performance of the controller. To overcome this problem, in this paper,a new output feedback PD (OPD) controller for position control of robot manipulatorsin task space is developed without velocity measurements.At last, taking into account the requirements of engineering a state feedbackPD-with-real-time-gravity-compensation control law with bounded input (SPD+) andan output feedback PD-with-real-time-gravity-compensation control law with bounderinput (OSPD+) are proposed. Both the two proposed control algorithms achieve goodcontrol performance.