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.

Generation of eigenstates using the phase-estimation algorithm

The phase estimation algorithm is so named because it allows an estimation of the eigenvalues associated with an operator. However, it has been proposed that the algorithm can also be used to generate eigenstates. Here we extend this proposal for small quantum systems, identifying the conditions under which the phase-estimation algorithm can successfully generate eigenstates. We then propose an implementation scheme based on an ion trap quantum computer. This scheme allows us to illustrate two simple examples, one in which the algorithm effectively generates eigenstates, and one in which it does not.

Historical PCDD Inputs and Their Source Implications from Dated Sediment Cores in Queensland (Australia)

Recent investigations have demonstrated the presence of an unidentified source of polychlorinated dibenzo-p-dioxins (PCDDs) in the coastal zone of Queensland (Australia). The present study provides new information on the possible PCDD sources and their temporal input to this environment. Two estuarine sediment cores were collected in northern Queensland for which radiochemical chronologies were established. Core sections from different depositional ages, up to three centuries, have been analyzed for 2,3,7,8-substituted PCDDs and polychlorinated dibenzofurans (PCDFs). Variations of PCDD concentrations in the sediment cores over several centuries of depositional history were relatively small, and elevated PCDD levels were still present in sediment slices from the early 17th century. PCDD/F isomer patterns and congener profiles in sediments deposited during the last 350 years were almost identical and correlated well to the characteristic profiles observed in surface sediments and soils from the entire Queensland coastline. Profiles were dominated by higher chlorinated PCDDs, in particular octachlorodibenzodioxin (OCDD), whereas PCDF concentrations were below or near the limit of detection. These results indicate the presence of a PCDD source prior to industrialization and production of commercial organochlorine products. Further, the present study demonstrates that PCDD input patterns have been similar along an extensive but localized area over at least several centuries, contributing relatively high concentrations of PCDDs to the coastal system of Queensland.

Evidence for the presence of a widespread PCDD source in coastal sediments and soils from Queensland, Australia

Recent studies have demonstrated the occurrence of elevated levels of higher chlorinated PCDDs in the coastal environment of Queensland, Australia. This study presents new data for OCDD contamination and full PCDD/F profile analysis in the environment of Queensland. Marine sediments, irrigation drain sediments and topsoil were collected from sites that were expected to be influenced by specific land-use types. High OCDD concentrations were associated mainly with sediments collected near the mouth of rivers which drain into large catchments in the tropical and subtropical regions. Further, analysis of sediments from irrigation drains could be clearly differentiated on the basis of OCDD contamination, with high concentrations in samples from sugarcane drains collected from coastal regions, and low concentrations in drain sediments from drier inland cotton growing areas. PCDD/F congener-specific analysis demonstrated almost identical congener profiles in all samples collected along the coastline. This indicates the source to be widespread. Profiles were dominated by higher chlorinated PCDDs, in particular OCDD whereas 2,3,7,8-substituted PCDFs were below the limit of quantification in the majority of samples. The full PCDD/F profile analysis of samples strongly resemble those reported for lake sediments from Mississippi and kaolinite samples from Germany, Strong similarities to these samples with respect to congener profiles and isomer patterns may indicate the presence of a similar source and/or formation process that is yet unidentified. (C) 2001 Elsevier Science Ltd. All rights reserved.

Optimal input states and feedback for interferometric phase estimation

We derive optimal N-photon two-mode input states for interferometric phase measurements. Under canonical measurements the phase variance scales as N-2 for these states, as compared to N-1 or N-1/2 for states considered bq previous authors. We prove, that it is not possible to realize the canonical measurement by counting photons in the outputs of the interferometer, even if an adjustable auxiliary phase shift is allowed in the interferometer. However. we introduce a feedback algorithm based on Bayesian inference to control this auxiliary phase shift. This makes the measurement close to a canonical one, with a phase variance scaling slightly above N-2. With no feedback, the best result (given that the phase to be measured is completely unknown) is a scaling of N-1. For optimal input states having up to four photons, our feedback scheme is the best possible one, but for higher photon numbers more complicated schemes perform marginally better.

Field test of a model of migration of moths (Lepidoptera : Noctuidae) in inland Australia

A migration of Helicoverpa punctigera (Wallengren), Heliothis punctifera (Walker) and Agrotis munda Walker was tracked from Cameron Corner (29degrees00'S, 141degrees00'E) in inland Australia to the Wilcannia region, approximately 400 km to the south-east. A relatively isolated source population was located using a distribution model to predict winter breeding, and confirmed by surveys using sweep netting for larvae. When a synoptic weather pattern likely to produce suitable conditions for migration developed, moths were trapped in the source region. The next morning a simulation model of migration using wind-field data generated by a numerical weather-prediction model was run. Surveys using sweep netting for larvae, trapping and flush counts were then conducted in and around the predicted moth fallout area, approximately 400 km to the south-east. Pollen carried on the probosces of moths caught in this area was compared with that on moths caught in the source area. The survey data and pollen comparisons provided evidence that migration had occurred, and that the migration model gave accurate estimation of the fallout region. The ecological and economic implications of such migrations are discussed.

Relationships between crude protein and mineral concentrations in alfalfa and value of alfalfa silage as a mineral source for periparturient cows

One hundred and eight samples from three cultivars of alfalfa were obtained from three cuttings in 1996-1998 to evaluate the relationship between crude protein (CP) and mineral concentrations of alfalfa with cutting and maturation. The CP content drastically decreased from 27.9 to 11.4% on DM with maturity. Highly positive correlations were observed between CP and K in the first and the second cutting of alfalfa. The Ca content remained almost constant throughout the growth period. Four multiparous Holstein cows were assigned an alfalfa silage diet or an orchardgrass silage diet from 3 weeks prepartum to 1 week postpartum to examine the effect on the mineral balance. In the prepartum and postpartum diet, the roughage to concentrate ratio was 70:30 and 50:50, with alfalfa being 50 and 100% of the roughage, respectively. The alfalfa contained 1.93% of K. No metabolic disorders occurred, but the body weight decreased drastically from 1 to 6 days postpartum with each diet because of the high milk production immediately after the parturition. Positive retention of N, Ca, P, Mg, and K was observed prepartum, whereas the cows had negative N and mineral retention from 2 to 4 days postpartum. The Ca and P absorption, and Mg retention of cows with the alfalfa diet were higher than with the grass diet. The plasma Ca and inorganic P were not affected by diet, but the plasma PTH at parturition and plasma hydroxyproline from 1 week prepartum to 1 week postpartum were higher with the alfalfa diet. These results suggest that the low K alfalfa is suitable not only to prevent the incidence of milk fever but also to increase Ca, P and Mg utilization of periparturient cows, but the mineral supplementation is needed for the postpartum cows immediately after the parturition. (C) 2001 Elsevier Science B.V. All rights reserved.

Sonographic estimation of fetal weight in macrosomic fetuses: diabetic versus non-diabetic pregnancies

The objective of this study is to compare the accuracy of sonographic estimation of fetal weight of macrosomic babies in diabetic vs non-diabetic pregnancies. Ali babies weighing 4000 g or more at birth, and who had ultrasound scans performed within one week of delivery were included in this retrospective study. Pregnancies with diabetes mellitus were compared to those without diabetes mellitus. The mean simple error (actual birthweight - estimated fetal weight); mean standardised absolute error (absolute value of simple error (g)/actual birthweight (kg)); and the percentage of estimated birthweight falling within 15% of the actual birthweight between the two groups were compared. There were 9516 deliveries during the study period. Of this total 1211 (12.7 %) babies weighed 4000 g or more. A total of 56 non-diabetic pregnancies and 19 diabetic pregnancies were compared. The average sonographic estimation of fetal weight in diabetic pregnancies was 8 % less than the actual birthweight, compared to 0.2 % in the non-diabetic group (p < 0.01). The estimated fetal weight was within 15% of the birthweight in 74 % of the diabetic pregnancies, compared to 93 % of the non-diabetic pregnancies (p < 0.05). In the diabetic group, 26.3 % of the birthweights were underestimated by more than 15 %, compared to 5.4 % in the non-diabetic group (p < 0.05). In conclusion, the prediction accuracy of fetal weight estimation using standard formulae in macrosomic fetuses is significantly worse in diabetic pregnancies compared to non-diabetic pregnancies. When sonographic fetal weight estimation is used to influence the mode of delivery for diabetic women, a more conservative cut-off needs to be considered.

Maximum Likelihood Estimation of Mixture Densities for Binned and Truncated Multivariate Data

Binning and truncation of data are common in data analysis and machine learning. This paper addresses the problem of fitting mixture densities to multivariate binned and truncated data. The EM approach proposed by McLachlan and Jones (Biometrics, 44: 2, 571-578, 1988) for the univariate case is generalized to multivariate measurements. The multivariate solution requires the evaluation of multidimensional integrals over each bin at each iteration of the EM procedure. Naive implementation of the procedure can lead to computationally inefficient results. To reduce the computational cost a number of straightforward numerical techniques are proposed. Results on simulated data indicate that the proposed methods can achieve significant computational gains with no loss in the accuracy of the final parameter estimates. Furthermore, experimental results suggest that with a sufficient number of bins and data points it is possible to estimate the true underlying density almost as well as if the data were not binned. The paper concludes with a brief description of an application of this approach to diagnosis of iron deficiency anemia, in the context of binned and truncated bivariate measurements of volume and hemoglobin concentration from an individual's red blood cells.

Impact of point source pollution on nitrogen isotope signatures (delta N-15) of vegetation in SE Brazil

This study presents novel evidence that N-15 natural abundance can be used as a robust indicator to detect pollutant nitrogen in natural plant communities. Vegetation from the heavily polluted industrial area of Cubatao in Sao Paulo State, SE Brazil, was strongly N-15 depleted compared to plants at remote sites. Historic herbarium samples from Cubatao were significantly less N-15 depleted than extant plants, indicating that N-15 depletion of vegetation is associated with present-day nitrogen pollution in Cubatao. The heavy load of nitrogenous atmospheric pollutants in Cubatao provides a nitrogen source for plants, and strongly N-15 depleted air NH3 is likely to contribute to plant and soil N-15 depletion. Epiphytic plants from Cubatao were extremely N-15 depleted (average -10.9parts per thousand) contrasting with epiphytes at remote sites (averages -1.0parts per thousand and -3.0parts per thousand). Nitrogen isotope composition of vegetation provides a tool to determine input of pollutant nitrogen into plant communities. The strong isotopic change of epiphytes suggests that epiphytes are particularly sensitive biomonitors for atmospheric pollutant nitrogen.

Establishment of Nilaparvata Iugens Stl in rice crop nurseries: a possible source of outbreaks

We tested the hypothesis that early-planted seedbeds of rioe are mere heavily infested with brown planthopper (BPH) than later seedbeds, and that transplanted plants with lBPH are a source of subsequent population increase and possible outbreaks. The experiments were conducted at CARDI and Takeo province in wet season 2000 and early wet 2 season 200 I. BPH at O. 25. 50, 100, 200 1m were infested onto plants with low and high fertilizer treatments. Rice seeds of varieties moderately and highly susceptible to BPH were sown 3 weeks early, 2 weeks early, at the normal time, and later than normal (5 weeks) and treated with low and high fertilizer rates. At Takeo, the 3< weeks early seedbeds were infested by BPH migration, and both varieties with high fertilizer caught more immigrant insects and subsequently had damaging outbreaks of BPH in the third generation. At CARDl, no seedbeds were infested with immigrant BPH. Seedbeds in areas with continuous cropping of rice have a high risk of BPH attack, Seedlings infested with 200, 100, and 50 BPI[/m2 resulted in death of the plant. Plants with 100 and 200 BPH/m'! were kj[Jed sooner. With 25 BPIVm2 plants were not kllled, but subsequent population increase caused yi eld reduction. Yield loss was high ill higlh fertilizer treated plants. Key words , ,