Hardware implementation of an artificial neural network using fieldprogrammable gate arrays (FPGA's)

In this paper, the authors present a hardware implementation of a fully digital multilayer perceptron artificial neural network using Xilinx Field Programmable Gate Arrays (FPGAs). Each node is implemented with two XC3042 FPGAs and a 1 K×8 EPROM. Training is done offline on a PC. The authors have tested successfully the performance of the network     

Contact control concepts in manipulation robotics-an overview

This paper presents the state of the art in the control of robotic manipulators in constrained motion tasks. Contact control concepts are classified using different criteria, and their main characteristics are analyzed. For each of the presented contact control concepts, the essential characteristics are stated. The paper covers some early ideas and their later improvements, as well as trends in this field. The advantages and drawbacks of the various control schemes are outlined, and they are compared from the standpoint of their implementation issues. In the paper, all characteristic results in the stability analysis of robotic manipulators in the constrained motion tasks are briefly reported. A new approach to the correct solution of contact tasks control is mentioned as well     

A boost-type converter for DC-supply of fluorescent lamps

A boost type converter is described that is suitable for low-voltage DC-supply of fluorescent lamps. It has inherent lamp current limitation (ballast action) and provides the high voltage pulses and electrode heating that are required for igniting the lamp. The proposed circuit is applicable in automotive, emergency, and portable light sources.<>     

A neural network control system with parallel adaptive enhancementsapplicable to nonlinear servomechanisms

In this paper, we present a technique for using an additional parallel neural network to provide adaptive enhancements to a basic fixed neural network-based nonlinear control system. This proposed parallel adaptive neural network control system is applicable to nonlinear dynamical systems of the type commonly encountered in many practical position control servomechanisms. Properties of the controller are discussed, and it is shown that if Gaussian radial basis function networks are used for the additional parallel neural network, uniformly stable adaptation is assured and the approximation error converges to zero asymptotically. In the paper, the effectiveness of the proposed parallel adaptive neural network control system is demonstrated in real-time implementation experiments for position control in a servomechanism with asymmetrical loading and changes in the load     

Enhancing the brightness of thin-film electroluminescent displaysby improving the emission process

AC thin film electroluminescent (ACTFEL) devices emit light laterally in addition to conventional surface emission. Because of the waveguiding in ACTFEL devices, lateral emission is stronger than surface emission. However, the former is ignored for displays. This note demonstrates that mirrors can be fabricated into substrates to reflect upward the lateral emission, which can then be summed with surface emission for displays. Use of standard silicon and thin film technology has demonstrated that the reflected lateral emission is a factor of three brighter than conventional surface emission. hence the effective efficiency of out-coupling the light has been improved for an ACTFEL device     

Enhanced high-frequency performance in a GaAs, self-aligned, n-JFETusing a carbon buried p-implant

C ion implantation has been employed, for the first time, to form the buried p-layer in GaAs, self-aligned, ion implanted JFETs. Comparable DC performance was seen for JFETs with C or Mg implants; however, C-backside JFETs showed superior high-frequency performance. High dose C-backside devices had a f_t of 28.3 GHz and a f_max of 43.2 GHz for a 0.5 μm gate length that were 28% and 46% higher, respectively, than comparable Mg-implanted JFETs. This enhancement is a result of the lower C_gs in the C-backside device resulting from he inherently low activation of the implanted C below the channel while the C still effectively compensated the tail of the Si-channel implant. This approach relaxes the trade-off between optimizing the DC and the AC performance for the buried p-implant in GaAs JFETs and MESFET's     

A digital automatic current control system based on an optimumassumed model

In this paper, a new ACR (automatic current regulator) scheme is introduced for brushless DC motors. It consists of a model closely resembling the motor and PWM inverter, an assumed rising pattern of the current, and a parameter estimation. Using the control scheme, the mean value of the motor current is kept at the reference current, and the transient current of the motor is kept equal to the rising pattern. For the disturbance, because of DC-line voltage change, the proportional path from the error to the duty cycle of the PWM signal stabilizes the ACR. This ACR scheme can be realized with a microcomputer     

A strategy for designing error detection schemes for general datanetworks

The author presents a five-step methodology that provides insight into the order in which errors should be considered when designing an error detection scheme, which types of error detection mechanisms are most effective, and which layer should be responsible for detecting a given type of error     

ATM switching and CPE adaptation in the North Carolina InformationHighway

This article describes the approach which has been used to implement a sector of the information highway throughout the state of North Carolina, with specific attention focused on the functionality of the equipment that comprises the North Carolina Information Highway (NCIH). This implementation of this information highway is based on ATM technology, used in both the switching systems of the network and the customer premises equipment (CPE). The large ATM switching systems provide the extensive number of connections that are predicted for the information highway. Multiple services can also be supported on the ATM switches. The information streams for these different services can all be converted into ATM format using the appropriate ATM adaptation layer for transport through the ATM network. By being able to support multiple services, and transport them across common ATM facilities, the ATM switches offer the NCIH great flexibility in the applications that can be provided to the customers. Use of the ATM technology allows the NCIH to provide the backbone for various services. To provide the distance learning service application and other applications on the NCIH, ATM capabilities are also being provided in CPE on a service multiplexer     

Models of visual word recognition: Sampling the state of the art.

A classification of models of visual word recognition is presented that facilitates formal comparisons between models of different formats. In light of the theoretical contributions to this special section, sets of criteria for the evaluation of models and strategies for model construction are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)