An adaptive neural-wavelet model for short term load forecasting

This paper proposed a novel model for short term load forecast in the competitive electricity market. The prior electricity demand data are treated as time series. The forecast model is based on wavelet multi-resolution decomposition by autocorrelation shell representation and neural networks (multilayer perceptrons, or MLPs) modeling of wavelet coefficients. To minimize the influence of noisy low level coefficients, we applied the practical Bayesian method Automatic Relevance Determination (ARD) model to choose the size of MLPs, which are then trained to provide forecasts. The individual wavelet domain forecasts are recombined to form the accurate overall forecast. The proposed method is tested using Queensland electricity demand data from the Australian National Electricity Market. (C) 2001 Elsevier Science B.V. All rights reserved.

Short-term electricity price forecasting using wavelet and SVM techniques

In deregulated electricity market, modeling and forecasting the spot price present a number of challenges. By applying wavelet and support vector machine techniques, a new time series model for short term electricity price forecasting has been developed in this paper. The model employs both historical price and other important information, such as load capacity and weather (temperature), to forecast the price of one or more time steps ahead. The developed model has been evaluated with the actual data from Australian National Electricity Market. The simulation results demonstrated that the forecast model is capable of forecasting the electricity price with a reasonable forecasting accuracy.

Effects of memory load and distraction on performance and event-related slow potentials in a visuospatial working memory task

Brain electrical activity related to working memory was recorded at 15 scalp electrodes during a visuospatial delayed response task. Participants (N = 18) touched the remembered position of a target on a computer screen after either a 1 or 8 sec delay. These memory trials were compared to sensory trials in which the target remained present throughout the delay and response periods. Distracter stimuli identical to the target were briefly presented during the delay on 30% of trials. Responses were less accurate in memory than sensory trials, especially after the long delay. During the delay slow potentials developed that were significantly more negative in memory than sensory trials. The difference between memory and sensory trials was greater at anterior than posterior electrodes. On trials with distracters, the slow potentials generated by memory trials showed further enhancement of negativity whereas there were minimal effects on accuracy of performance. The results provide evidence that engagement of visuospatial working memory generates slow wave negativity with a timing and distribution consistent with frontal activation. Enhanced brain activity associated with working memory is required to maintain performance in the presence of distraction. © 1997 by the Massachusetts Institute of Technology

Titanium phosphate for Fuel Cell Proton Conduction Membranes

Environmental issues due to increases in emissions of air pollutants and greenhouse gases are driving the development of clean energy delivery technologies such as fuel cells. Low temperature Proton Exchange Membrane Fuel Cells (PEMFC) use hydrogen as a fuel and their only emission is water. While significant advances have been made in recent years, a major limitation of the current technology is the cost and materials limitations of the proton conduction membrane. The proton exchange membrane performs three critical functions in the PEMFC membrane electrode assembly (MEA): (i) conduction of protons with minimal resistance from the anode (where they are generated from hydrogen) to the cathode (where they combine with oxygen and electrons, from the external circuit or load), (ii) providing electrical insulation between the anode and cathode to prevent shorting, and (iii) providing a gas impermeable barrier to prevent mixing of the fuel (hydrogen) and oxidant. The PFSA (perfluorosulphonic acid) family of membranes is currently the best developed proton conduction membrane commercially available, but these materials are limited to operation below 100oC (typically 80oC, or lower) due to the thermochemical limitations of this polymer. For both mobile and stationary applications, fuel cell companies require more durable, cost effective membrane technologies capable of delivering enhanced performance at higher temperatures (typically 120oC, or higher. This is driving research into a wide range of novel organic and inorganic materials with the potential to be good proton conductors and form coherent membranes. There are several research efforts recently reported in the literature employing inorganic nanomaterials. These include functionalised silica phosphates [1,2], fullerene [3] titania phosphates [4], zirconium pyrophosphate [5]. This work addresses the functionalisation of titania particles with phosphoric acid. Proton conductivity measurements are given together with structural properties.

A framework for electricity price spike analysis with advanced data mining methods

There are many techniques for electricity market price forecasting. However, most of them are designed for expected price analysis rather than price spike forecasting. An effective method of predicting the occurrence of spikes has not yet been observed in the literature so far. In this paper, a data mining based approach is presented to give a reliable forecast of the occurrence of price spikes. Combined with the spike value prediction techniques developed by the same authors, the proposed approach aims at providing a comprehensive tool for price spike forecasting. In this paper, feature selection techniques are firstly described to identify the attributes relevant to the occurrence of spikes. A simple introduction to the classification techniques is given for completeness. Two algorithms: support vector machine and probability classifier are chosen to be the spike occurrence predictors and are discussed in details. Realistic market data are used to test the proposed model with promising results.

Electrical impedance particle size method (Coulter Counter) detects the large fat globules in poorly homogenised UHT processed milk

The large fat globules that can be present in UHT milk due to inadequate homogenisation cause a cream layer to form that limits the shelf life of UHT milk. Four different particle size measurement techniques were used to measure the size of fat globules in poorly homogenised UHT milk processed in a UHT pilot plant. The thickness of the cream layer that formed during storage was negatively correlated with homogenisation pressure. It was positively correlated with the mass mean diameter and the percentage volume of particles between 1.5 and 2 mu m diameter, as determined by laser light scattering using the Malvern Mastersizer. Also, the thickness of the cream layer was positively correlated with the volume mode diameter and the percentage volume of particles between 1.5 and 2 mu m diameter, as determined by electrical impedance using the Coulter Counter. The cream layer thickness did not correlate significantly with the Coulter Counter measurements of volume mean diameter, or volume percentages of particles between 2 and 5 mu m or 5 and 10 mu m diameter. Spectroturbidimetry (Emulsion Quality Analyser) and light microscopy analyses were found to be unsuitable for assessing the size of the fat particles. This study suggests that the fat globule size distribution as determined by the electrical impedance method (Coulter Counter) is the most useful for determining the efficiency of homogenisation and therefore for predicting the stability of the fat emulsion in UHT milk during storage.