In this paper, a neural-network-based guidance methodology that utilizes line-of-sight based task-space sensory feedback is proposed for the localization of autonomous robotic vehicles. The novelty of the overall system is its applicability to cases that do not allow for the direct proximity measurement of the vehicle's pose (position and orientation). Herein, the proposed neural-network (NN) based guidance methodology is implemented on-line during the final stage of the vehicle's motion (i.e., docking). The systematic motion errors of the vehicle are reduced iteratively by executing the corrective motion commands, generated by the NN, until the vehicle achieves its desired pose within random noise limits. The guidance methodology developed was successfully tested via simulations for a 6-dof (degree-of-freedom) vehicle and via experiments for a 3-dof high-precision planar platform

A novel rendezvous-guidance algorithm is proposed in this paper for the robotic interception of moving objects. The object is assumed to be moving on an a priori unknown trajectory, though, its instantaneous state is visually observed and fed to the proposed algorithm. The output of the guidance algorithm is an instantaneous end-effector velocity command with the largest possible closing velocity component while not violating the rendezvous criteria and the dynamic limitations of the actuators of the robot. The proposed method reduces the interception-times achieved via conventional tracking methods as demonstrated by simulation examples.

An Ideal Proportional Navigation Guidance (IPNG) technique is applied to the problem of intercepting a fast maneuvering object in real time using an industrial robotic manipulator. A hybrid method is utilized, in which IPNG is combined with a conventional tracking method to ensure smooth interception. The method is suited for the interception of both fast maneuvering as well as slow maneuvering targets, and does not require long term target trajectory prediction. Experimental results are reported.

Presents an approach to online, robot-motion planning for moving-object interception. The proposed approach utilizes a navigation-guidance-based technique, that is robust and computationally efficient for the interception of fast-maneuvering objects. Navigation-based techniques were originally developed for the control of missiles tracking free-flying targets. Unlike a missile, however, the end-effector of a robotic arm is connected to the ground, via a number of links and joints, subject to kinematic and dynamic constraints. Also, unlike a missile, the velocity of the robot and the moving object must be matched for a smooth grasp, thus, a hybrid interception scheme, which combines a navigation-based interception technique with a conventional trajectory tracking method is proposed herein for intercepting fast-maneuvering objects. The implementation of the proposed technique is illustrated via numerous simulation examples.

This paper presents a novel approach to online robot-motion planning for moving-object interception. A hybrid scheme, which combines a navigation-based technique with a conventional trajectory tracking method, is proposed for intercepting fast-manoeuvring objects. The implementation of the proposed technique is discussed via a simulation example.

A number of active prediction planning and execution (APPE) systems have recently been proposed for robotic interception of moving objects. The cornerstone of such systems is the selection of a robot-object rendezvous-point on the predicted object trajectory. Unlike tracking-based systems, which minimize the state difference between the object and the robot at each control period, in this methodology the robot is sent directly to the selected rendezvous-point. A fine-motion tracking strategy would then be employed for grasping the moving object. Herein, a novel strategy for selecting the optimal (earliest) rendezvous-point is presented. For objects with predictable trajectories, this is a significant improvement over previous APPE strategies which select the rendezvous-point from a limited number of non-optimally chosen candidates.

The remote manipulator described consists of two arms: a 3-DOF main lifting arm (aerial device), and a 6-DOF robot manipulator (treejib) mounted at the upper end of the aerial device. Each arm is single-joint rate controlled, with the reference joint rate being proportional to the corresponding joint displacement of a corresponding hand controller. Both controllers are kinematically equivalent to each of the two arms. Two of the treejib joints are instrumented for position feedback, and the corresponding hand controller joints are motorized and also position controlled. This ensures that no excessive force is exerted on the tool during contact with the branches. The system is of a master-slave rate controlled force reflective type. The design presented has novel features such as a quick connect-disconnect coupler for hydraulic tools and a force reflective master in a form of a small kinematic replica of the slave. The system has been extensively field tested. The results were very promising, however several modifications to the basic concept used to automate the specific operations are necessary.>

Point-to-point quasi-static motion planning is essential for effectively controlling the elastic deformations and vibrations of the end-effector of flexible-link manipulators. In this paper, this problem is formulated considering the kinematic equations and link deflection equations of the manipulator. An algorithm is proposed to obtain the numerical solution to the problem by utilizing the Newton-Raphson method. The kinematic equations and link deflection equations, as well as their differential relationships, are derived by using the algebra of rotations. The effectiveness of the proposed algorithm is demonstrated in examples.>

In this paper a general method is presented, based on the data of three point positions, velocities and accelerations of the end effector, for solving the forward kinematics problem of any platform-type manipulator, including the 6 DOF Stewart Platform. Numerical examples are included to demonstrate the application of the method. It is shown that the equations for the forward position kinematics are highly-nonlinear, however, closed-form solutions to the forward rate kinematics and the forward acceleration kinematics can be obtained by solving a system of linear equations. The advantages of using additional passive joint encoders is also discussed, to simplify the solution of the position kinematics problem, and obtain a one-to-one relation between the actuated joint variables and the end effector configurations.>

A spacecraft-manipulator system is considered in this paper. Dynamic singularities are the singularities occurring when inverting the system generalized Jacobian required by the conventional method for solving the inverse kinematics of space manipulators. To avoid dynamic singularities three methods are developed here based on the manipulator Jacobian, instead of the system generalized Jacobian. These methods are compared with the conventional method on the basis of their convergence rate, accuracy and sensitivity. Results of this comparison are presented in this paper.>

This paper investigates the kinematic singularities of a general mechanism (arbitrary kinematic chain), considered as a non-redundant input-output device with equal number of inputs and outputs. The instantaneous kinematics of a mechanism is described by the orientation of a linear subspace, the motion space, inside the velocity space of all potential instantaneous motions. The definition of singularity for a general mechanism is provided. On the basis of the six types of singular configurations introduced in the paper a comprehensive singularity classification is developed.>

This paper presents a new generalized approach to the singularity analysis of a general mechanism (arbitrary kinematic chain), considered as a non-redundant input-output device with equal number of inputs and outputs. The instantaneous kinematics of a mechanism is described by means of a velocity equation, explicitly including not only the input and output velocities but also the passive-joint velocities. A precise definition of singularity of a general mechanism is provided. On the basis of the six types of singular configurations introduced in the paper singularity classifications are proposed.>

The basic quantities required for computations in robot kinematics, dynamics, and control are defined. By modifying the rotational operator in terms of the Euler parameters and appropriately choosing the zero reference position, the method of the algebra of rotations becomes more efficient than the homogeneous transformation method for computing these basic quantities. As a result, the computational efficiency is improved by about 50|% for robot kinematics, and by about 20|% for robot dynamics and control, if the method of the algebra of rotations is used instead of the homogeneous transformation method.>

In this paper, a complete model of the elasto-kinematics is formulated in terms of a new kinematic notation, called the algebra of rotations. Based on this formulation, the elegant and concise expressions are derived for the displacement equation and especially the Jacobians governing the motion mapping between the manipulator tip and joint variables and link deflections. Introduction of the elasto-kinematics into the elasto-dynamics can directly take into consideration the nonlinear coupling between joint variables and link deflections, and thus improve the result of the elasto-dynamics.>

A remote manipulator was developed for forestry operations to relocate operators away from hazardous operations. The operations of tree trimming are performed in the vicinity of live electrical distribution lines, generating a hazard for the operator. The remote manipulator has two major subsystems: a three degree-of-freedom (DOF) aerial device acting as the main lifting arm, and a six DOF manipulator mounted at the end of the aerial device. Each arm was rate controlled with the reference rate provided by the displacement of a corresponding hand controller, which was kinematically equivalent to the manipulator. The mechanical design, the control system of the remote manipulator, and the results of field tests are described.>

An algorithm for motion planning of flexible manipulators in quasi-static operations is proposed. This method is based on the algebra of rotations. A concise motion expression for flexible manipulators is developed to reflect the contributions of joint motions and link deflections to the motion of the end effector by three respective Jacobians. The proposed algorithm is composed of two major loops: the first for joint motions and the second for link deflections. The problems associated with the implementation of the algorithm are investigated, including convergency analysis, singularity analysis, link deflection analysis, etc. It is found that the algorithm is efficient and accurate for motion planning of flexible link manipulators.>

Joint rate solutions for redundant manipulators are formulated in terms of a particular solution and a homogeneous solution formed using a basis for the Jacobian null space. Solutions for several local optimization objectives are formulated in terms of Jacobian null-space bases. A method based on a decomposition of screw coordinates is presented for finding null-space bases and particular solutions. The screw decomposition is applied in the derivation of the analytical expressions for the null-space bases and in the identification of special configurations for a seven-revolute manipulator. The analysis is then extended to multiple-arm systems with common degrees of freedom, and is applied to an analytical example involving industrial manipulators mounted on a common platform.>

Various definitions of task-space tolerances, using geometric pseudoenvelopes, are introduced, and new approaches to "direct" and "inverse" joint-error analysis of robots are formulated. These error analyses lead to the development of a "feasible joint-tolerance domain" concept for use in computer-aided design of robots.

The kinematic transformation between task space and joint configuration coordinates is nonlinear and configuration dependent. A solution to the inverse kinematics is a vector of joint configuration coordinates that corresponds to a set of task space coordinates. For a class of robots closed form solutions always exist, but constraints on joint displacements cannot be systematically incorporated in the process of obtaining a solution. An iterative solution is presented that is suitable for any class of robots having rotary or prismatic joints, with any arbitrary number of degrees of freedom, including both standard and kinematically redundant robots. The solution can be obtained subject to specified constraints and based on certain performance criteria. The solution is based on a new rapidly convergent constrained nonlinear optimization algorithm which uses a modified Newton-Raphson technique for solving a system nonlinear equations. The algorithm is illustrated using as an example a kinematically redundant robot.