Process monitoring using auto-associative,feed-forward artificial neural networks

The potential of using artificially simulated neural networks as intelligent, adaptive process-monitoring devices is discussed. The investigation is considered as a method for automatic, intelligent exception reporting for quality control applications. The technique is also compared with the conventional statistical approaches of principal component analysis and Kohonen's feature map. The applications of the technique in aerospace and manufacturing environments are presented and a possible extension of the method to incorporate a diagnostic function is discussed.Seconded from Cheltenham and Gloucester College of Higher Education as a Royal Society/SERC Research Fellow at Smith's Industries Aerospace and Defence Systems, Bishop's Cleeve, Cheltenham, UK.     

Efficient simulation of hierarchical stochastic activity network models

Stochastic extensions to Petri nets have gained widespread acceptance as a method for describing the dynamic behavior of discrete-event systems. Both simulation and analytic methods have been proposed to solve such models. This paper describes a set of efficient procedures for simulating models that are represented as stochastic activity networks (SANs, a variant of stochastic Petri nets) and composed SAN-based reward models (SBRMs). Composed SBRMs are a hierarchical representation for SANs, in which individual SAN models can be replicated and joined together with other models, in an iterative fashion. The procedures exploit the hierarchical structure and symmetries introduced by the replicate operation in a composed SBRM to reduce the cost of future event list management. The procedures have been implemented as part of a larger performance-dependability modeling package known asUltraSAN, and have been applied to real, large-scale applications. This work was supported in part by the Digital Equipment Corporation Faculty Program: Incentives for Excellence.     

On the cover.

Discusses the cover art for this issue of American Psychologist. Portions of the article are based on an interview with the artist, Melissa Miller, on November 6, 2006. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Multistage Zeolite Drying for Energy-Efficient Drying

This work discusses the potential of three multistage zeolite drying systems (counter-, co-, and cross-current) with a varying number of stages. The evaluation showed that for 2-4 stages with heat recovery the efficiency of the systems ranges between 80 and 90%. Additionally, by introducing a compressor, the latent heat in the exhaust air from the regenerator is recovered and used to heat the inlet air for an additional drying stage. As a result, for the counter-current drying system and compressor pressure 1.5-2 bar, a maximum energy efficiency of 120% is achieved, which results in halving the energy consumption compared to conventional drying systems.     

A quantitative approach to modeling mammalian myelinated nervefibers for electrical prosthesis design

Presents an upgraded cable model of mammalian myelinated nerve fibers in an extracellularly applied field. The kinetics of the nodes is based upon voltage clamp data in rat motor fibers at 37°C (J.R. Schwartz and G. Eikhof, 1987), while the resting membrane potential is computed with the Goldman equation. The resulting spike shape, conduction velocity, strength/duration behavior, and absolute and relative refractory period are in good quantitative agreement with published experimental data in mammals at normal body temperature and at 20°C. Results at intermediate temperatures however, suggest that the widely used concept of a constant Q₁₀ for the rate constants is invalid. In addition, the model generates realistic abortive spikes towards the end of the absolute refractory period and it can describe the consequences of repetitive firing. The results stress the advantages of a multiple nonlinear node model even if only time aspects of nerve behavior are under study. It turned out, that the model presented here describes in vivo neural properties relevant for electrical prosthesis design better than previous models in literature     

High-risk sexual behavior at a midwestern university: a confirmatory survey

According to a 1991 study of sexual behavior based on a random sample of heterosexual undergraduates at a Midwestern university, 80% of the males and 73% of the females had experienced vaginal or anal intercourse. The average age at first vaginal intercourse was 17.2 years for both sexes. Seventeen percent of the sexually experienced males and 18% of the sexually experienced females had engaged in heterosexual anal intercourse; among these respondents, the average age at first anal intercourse was 20.3 for males and 19.1 for females. Although less than four years, on average, had elapsed since the respondents had first had vaginal intercourse, males reported an average of 8.0 lifetime vaginal-sex partners and females reported an average of 6.1. Overall, the findings from this random sample of students are similar to those from a 1988 convenience sample of the same college population.     

When 3-year-olds pass the appearance–reality test.

68 3-yr-olds received a standard appearance–reality task along with either a trick task, in which the appearance question was placed in the context of a deceptive game, or a reduced information processing task, in which a dual object (e.g., a sponge-rock) was presented along with an object that matched the dual object's identity (a sponge) and one that matched the dual object's appearance (a rock). Children were more likely to pass either the trick or reduced information processing task and fail the standard than the reverse. Thus, 3-yr-olds can grasp the distinction between appearance and reality (a) when their goal is to trick someone, which may prime them to think about the other's mental state, and (b) when they do not need to hold conflicting object identities in mind at the same time. (PsycINFO Database Record (c) 2010 APA, all rights reserved)