Throughput Analysis for a High-Performance FPGA-Accelerated Real-Time Search Application

We propose an FPGA design for the relevancy computation part of a high-throughput real-time search application. The application matches terms in a stream of documents against a static profile, held in off-chip memory. We present a mathematical analysis of the throughput of the application and apply it to the problem of scaling the Bloom filter used to discard nonmatches.

Roadway-Vehicle Interaction, Physical Indexes, and Human Judgment of Ride Quality

Present serviceability index (PSI) modeling has been an important subject for decades. Other dynamic indexes characterizing a roadway such as the international roughness index (IRI), averaged rectified slope (ARS), and averaged rectified speed (ARV) have been proposed and studied. However, the roles played by these indexes in the interaction between road, vehicle, and human ratings have not been made clear. A unified physical model linking the static profile of a roadway and the dynamic response of a vehicle to the profile to the serviceability index of the roadway is presented here. Analytical expressions for jerk index, acceleration index, ARV, ARS, and IRI are derived from the developed model in terms of the physical parameters for roadways and the dynamic characteristics of a vehicle. Then a linear relation between the PSI and the logarithm of the jerk index is proposed. Using the jerk index computed from field profile data, the linear functional form for the PSI is verified, and regression R2 values higher than 0.94 are obtained for various types of pavements. The same analysis is performed for other dynamic indexes, and the R2 values are found to be approximately in the range from 0.70 to 0.80. These results indicate that the theoretical model correctly predicts and explains the human rating of ride quality and that the jerk experienced by raters in a moving vehicle dictates the ratings. Moreover, the relationship of the static parameters of roadway profiles with human ratings is discussed using the conventional approach.

Tension of Conductor Under Concentrated Loads

A common difficulty in the analysis and design of transmission and distribution lines is to determine a conductor’s tension and its static profile under concentrated loads. For relatively small concentrated loads (such as detuning pendulums on transmission lines), approximation methods may give good predictions. For large concentrated loads (such as fallen trees on distribution lines), however, exact solutions must be found. This paper presents methodologies to compute conductor tension and static profile in three-dimensional space using both approximate and exact solution procedures under concentrated loads with different boundary conditions. Practical engineering examples from galloping control of transmission lines and mechanical coordination of distribution lines are given to demonstrate the applicability of the theory.

Surface Profile Measurement During Turning Using Fringe-Field Capacitive Profilometry

The use of fringe-field capacitive sensing for surface profile measurement during the turning process is described. Measurements of the local surface height are inferred from variations in a fringe electric field induced between the sensing element and the workpiece. The surface profile is determined from high-speed scanning of the sensing element across the surface. The technique is particularly well-suited to the relatively harsh environment of in-process measurement. We have implemented a system in which profile measurements are made continuously, in real-time, and immediately adjacent to the cutting tool. The results of tests conducted to determine the accuracy and sensitivity of this capacitive profilometer are presented. In-process measurements of surfaces generated by turning with roughness in the range of 0.3 to 4.0 μm were made. Comparisons with static profile measurements made using standard stylus instrumentation are presented, and show quantitative agreement.