An implementation of a real-time scheme suitable for coordinated-motion control of a class of teleoperated industrial machines with redundancy is presented. An efficient gradient projection technique is adopted for the numerical solution. The scheme utilizes the redundancy to avoid the joint limits by minimizing a hyperbolic function of the joint distances from the mid-range. This new performance criterion is shown to be advantageous over similar criteria; both the joint-limit avoidance capability and the resulting joint velocity profiles can be adjusted by the appropriate choice of parameters introduced in the criterion. Previous criteria are shown to be special cases of the new criterion. Furthermore, the scheme includes a novel algorithm which incorporates the bounded joint velocities. The joint motions are determined considering both the required task in terms of the desired end-effector speed and the dynamic considerations, such as hydraulic circuit interdependency and power limitations.

The feasibility and effectiveness of the implementation of the scheme is first tested through simulations of a Kaiser Spyder-like excavator machine on a PC-486 micro-computer. The robustness and real-time response of the scheme are then validated on a real-time excavator-based graphics simulator interfaced to a human operator through a joystick.

This paper discusses three methods of training multi-layer perceptrons (MLPs) to model a six-degrees-of- freedom inertial sensor. Such a sensor is designed for use with a robot to determine the location of objects it has to pick up. The sensor operates by measuring parameters related to the inertia of an object and computing its location from those parameters. MLP models are employed for part of the computation. They are trained to output the orientation of the object in response to an input pattern that includes the period of natural vibration of the sensor on which the object rests. After reviewing the working principle of the sensor, the paper describes the three MLP training methods (backpropagation, optimisation using the Levenberg-Marquardt algorithm, evolution based on the genetic algorithm) and presents the experimental results obtained.

The algorithms of inverse kinematics based on optimality constraints have some problems because those are based only on necessary conditions for optimality. One of the problems is a switching problem, i.e., an undesirable configuration change from a maximum value of a performance measure to a minimum value may occur and cause an inverse kinematic solution to be unstable. In this paper, we derive sufficient conditions for the optimal solution of the kinematic control of a redundant manipulator. In particular, we obtain the explicit forms of the switching condition for the optimality constraintsbased methods. We also show that the configuration at which switching occurs is equivalent to an algorithmic singularity in the extended Jacobian method. Through a numerical example of a cyclic task, we show the problems of the optimality constraints-based methods. To obtain good configurations without switching and kinematical singularities, we propose a simple algorithm of inverse kinematics.

A light-weight, high-torque actuator with accurate torque control capability is described. The actuator uses a small hydrostatic transmission to achieve the advantage of large gear reduction from a high speed DC motor, and retains accurate joint torque sensing and control capabilities with no backlash. A disadvantage of the actuator is that is introduces extra dynamics which must be accounted for in robot control systems. It is shown that state feedback enables closed loop control of joint torque, with full back drivability, through an effective gear ratio of 485:1 for the experimental system. The actuator can therefore be used for both position control and output force control, which is essential for modern robot control algorithms. A mathematical model of the system is presented in this paper along with experimental results.

In the tracking control of manipulators via the sliding scheme, it may happen that sometimes, because of various inaccuracies, the definition of the actual sliding surface involves errors terms which may be either deterministic or, on the contrary, stochastic. This paper considers this last case and shows how one can estimate the new performances of the system so disturbed. A stochastic Hamilton's principle is applied, by combining the Lagrange parameter technique with results of the dynamic programming approach.