G. Thomas

Leveraging smart grid technology for home health care

Smart grid technology is emerging as a powerful method for managing home energy consumption and improving the efficiency of power delivery. Due to declining birth rates and advances in health care, the world population is aging and new technology can assist in care of elderly populations. Smart grid technology can be used to provide useful information on the activities of daily living and can be used to monitor both the short-term and long-term health of elderly individuals. This report outlines some experiments that demonstrate the concept.

Using a wireless LAN to perform motion detection

In a wireless home network, multipath signal components are made up of reflections from walls, ceilings, floors, furniture, and other objects. Any motion in the vicinity will change the magnitude of the multipath components arriving at the antenna. This variation in signal strength intensity is statistically significant and does not rely on line-of-sight between the transmitter and receiver. In this project, a set of wireless sensor nodes were developed that are capable of utilizing the RSSI variance between path links to detect motion and control electrical outlets. The result is a stable, reliable room of “smart outlets” that can sense whether or not a room is occupied, and change their state accordingly.

Integration of process modeling with laboratory experiments in conceptual design: Bio-based glycerol dehydration case study

Abstract During the conceptual stage of process design, the experimental work to determine the reaction parameters is often carried out independently from the simulation work used to develop the conceptual process. The result of this disconnect is that the optimal process may remain undiscovered, since the process design engineer is constrained by the reaction conditions originally studied in the laboratory. By utilizing a methodology for integrating the process development with the laboratory experiments at the earliest stages, the process designer can ensure that the laboratory data is gathered only for economically viable and technically feasible process conditions. The methodology is illustrated through a case study on dehydration of bio-based glycerol.

Separation of merged mass spectral patterns by feed-forward neural network filtering

This paper describes the separation of merged signals from a mass-selective chromatographic detector by means of an adaptive filtering technique. The technique is based on parallel feed-forward neural networks, which are trained to resolve the mass spectra of two merged chemical compounds. Specifically, the chemical mass spectra of the compounds ethyl benzene and xylene were used to evaluate a filter based on probabilistic neural networks (PNN). The results are that the PNN filter shows good noise rejection and is fast enough computationally to be utilized in real time. The filter technique has applications in on-line processing of environmental monitoring instrumentation data and direct processing of pixel spectral data, such as hyperspectral image cubes.