Three nonlinear continuous-time predictive control schemes are proposed to address the trajectory tracking control problem of rigid link robot manipulators. The control laws using state variable feedback minimize a quadratic performance index of the state predicted tracking error. Without online optimization, an asymptotic tracking of smooth reference trajectories is guaranteed. The proposed controllers achieve the positions and speed tracking objectives via link position measurements. The Lyapunov theory is used to prove the boundedness and stability convergence of the state tracking. Robustness with respect to payload uncertainties and viscous friction is shown. Simulations for a two-link rigid robot are performed to validate the proposed controller.

No abstract

No abstract

No abstract

No abstract

No abstract

The experimental robotic system, given by Architecture of the RobuTER/ULM mobile manipulatorRobuTER/ULM is composed of a rectangular differentially-driven mobile base on which is mounted a manipulator. The robot is controlled by an on-board MMX industrial PC and by four MPC555 microcontroller cards communicating via a CAN bus. The on-board PC is running under Linux 6.2 with RTAI layer 1.3. This layer interfaces C/C++ application with that developed under SynDEx (http://www.syndex.org). The first MPC555 card controls the mobile base. The second and the third control the first three and the last three joints of the manipulator. The last MPC555 controls the effort sensor. The mobile base has two driven wheels ensuring its mobility and two free wheels to maintain its stability. The mobile base is equipped with a belt of 24 ultrasonic sensors, a laser measurement system at the front and an odometer sensor on each driven wheel. www.intechopen.com Robot Manipulators, New Achievements 148The manipulator is a six-dof ultra-light manipulator (ULM) with two-finger electrical gripper. All of the joints are rotatable. The manipulator is equipped with incremental position sensor for each articulation and with a six-dof effort sensor integrated on the gripper. The robot is also equipped with a monochrome CCD camera placed on the gripper (eye-inhand camera) with an acquisition card. The resolution of the camera is 352*240 pixels. Images are directly transmitted to the off-board PC via the wireless video transmission system. The camera is maneuverable enough to explore the environment of the robot due to the six dof of the manipulator. Kinematic analysis of RobuTER/ULM Main reference framesThe kinematic analysis of the robot needs to focus on the following main reference frames and transformation matrices (Fig. 2) Kinematic analysis of the ULM manipulatorThe position coordinates and orientation angles of the end-effector are calculated in R M by (1) following the Modified Denavit-Hartenberg (MDH) representation (Khalil & Kleinfinger, 1986) whereM T 2 , 6 T E and k-1 T k are given by (2) (Dombre & Khalil, 2007): Kinematic analysis of the mobile manipulatorIt involves the interaction between the mobile base and the manipulator. The location of the end-effector is given in R A by:A T B and B T M are given by (4) and (5) For RobuTER/ULM, as shown in Fig. 3, Hentout et al., 2009a).www.intechopen.com . Each agent models a principal function of the mobile manipulator and manages a different sub-system. In addition, for each agent corresponds a mechanism connecting the four capacities Supervision, Perception, Decision and Action explained in more details in (Hentout et al., 2008).The Supervision capacity is a virtual entity that select modules which result in the necessary behaviour facing a given situation.The following are the basic functions of the architecture agents (Hentout et al., 2009b):  SA, Supervisory Agent: SA receives the mission to be executed, decides on its feasibility according to the status and the availability (P...

Force/position control strategies provide an effective framework to deal with tasks involving interaction with the environment. One of these strategies proposed in the literature is external force feedback loop control. It fully employs the available sensor measurements by operating the control action in a full dimensional space without using selection matrices. The performance of this control strategy is affected by uncertainties in both the robot dynamic model and environment stiffness. The purpose of this paper is to improve controller robustness by applying a neural network technique in order to compensate the effect of uncertainties in the robot model. We show that this control strategy is robust with respect to payload uncertainties, position and environment stiffness, and dry and viscous friction. Simulation results for a three degrees-of-freedom manipulator and various types of environments and trajectories show the effectiveness of the suggested approach compared with classical external force feedback loop structures.