Dynamic modelling using neural networks

A neural network structure is presented that uses feedback of unmeasured system states to represent dynamic systems more efficiently than conventional feedforward and recurrent networks, leading to better predictions, reduced training requirement and more reliable extrapolation. The structure identifies the actual system states based on imperfect knowledge of the initial state, which is available in many practical systems, and is therefore applicable only to such systems. It also enables a natural integration of any available partial state-space model directly into the prediction scheme, to achieve further performance improvement. Simulation examples of three varied dynamic systems illustrate the various options and advantages offered by the state-feedback neural structure. Although the advantages of the proposed structure, compared with the conventional feedforward and recurrent networks, should hold for most practical dynamic systems, artificial systems can readily be created and real systems can surely be found for which one or more of these advantages would vanish or even get reversed. Caution is therefore recommended against interpreting the suggested advantages as strict theorems valid in all situations.     

Enterobacter sakazakii survives spray drying

Four strains of Enterobacter sakazakii were inoculated into 35% reconstituted skim milk at 10⁷ and 10² cfu/g dry wt. After spray-drying in a Buchi mini spray drier (inlet 160°C and outlet 90°C), the resulting powders were analysed for E. sakazakii by enrichment and enumeration. In all cases, E. sakazakii survived the spray drying process and were detected in the powders with a low inoculum and enumerated in all the powders with the high inoculum for at least 12 weeks. The results emphasize that the controls in place to prevent E. sakazakii from getting to the spray drier are essential.     

Effects of Electrical Stimulation on the Functional Properties of Lamb Muscle

One side from each of eleven lamb carcasses was electrically stimulated (60 Herz alternating current at 240v followed by 420v) within 45 min postmortem. Three carcasses (stimulated and respective control sides) were rapidly chilled (1°C), three carcasses received a delayed chill (12°C until pH 6.0 was attained, then excision and 1°C), and three carcasses were slowly chilled (12°C). Measurements beginning at 44 hr postmortem showed that stimulation increased the sarcomere lengths but had few consistent or significant effects on water-holding capacity, protein solubility, emulsifying capacity, gel strength, cooking loss, or binding strength. The remaining two carcasses were rapidly chilled and their meat was manufactured into frankfurters. The emulsion stability was slightly improved by stimulation but smokehouse weight loss, Warner-Bratzler shear force, penetration force, cooking loss, and sensory characteristics were unaffected.     

Effect of Sodium Chloride Levels in Frankfurters on the Growth of Clostridium sporogenes and Staphylococcus aureus

Frankfurters were made with 100%, 80%, and 60% of the normal 2.5% added NaCl, plus 150 ppm NaNO₂ and 430 ppm ascorbate. They were either untreated or inoculated with Staphylococcus aureus or Clostridium sporogenes and incubated at a moderate abuse temperature of 11°C for up to 11 days. A slightly more rapid growth of the natural flora was noted in the frankfurters containing 60% of the normal salt compared to those containing higher levels of salt. No outgrowth of clostridial spores occurred in any inoculated sample. Staphylococcal growth was slow and inconsistent, with no effect related to salt levels. Incubating frankfurters at higher tern. peratures showed that abuse temperature had a greater effect on total aerobic and staphylococcal growth than did the salt levels tested.     

The System K2SO4-Cs2SO4

The phase diagram for the system K₂SO₄-Cs₂SO₄ was determined by using DTA for melting relations and DTA and high-temperature X-ray diffractometry for subsolidus relations. At the solidus the system shows complete solid solubility, with a minimum at 940°C and 50 mol% Cs₂SO₄. Orthorhombic K₂SO₄ and Cs₂SO₄, the stable low-temperature forms, show mutual solid solubility and form a eutectoid at 50 mol% Cs₂SO₄ and 430°C, the lowest temperature of stability of the high-temperature hexagonal solid-solution phase. Isothermal plots of the a and c dimensions of this hexagonal phase vs composition show large positive deviations from linearity for c. These deviations are interpreted on the basis of the crystal structure of KNaSO₄ with a similar unit cell.     

Damage to Pork Liver Caused by Repeated Freeze-Thaw Cycling and Refrigerated Storage

The effects of freezing and thawing (F-T) and refrigerated storage (R-S) on pork liver were examined. Intact cells were isolated from the tissue and analyzed for membrane integrity and the tissue extracts were analyzed for protein content. Significant (p < 0.05) differences were noted in the number of isolated intact cells and in tissue protein content between F-T and R-S or fresh livers. Differences in cell membrane integrity were also noted between fresh and R-S livers. Repeated freezing and thawing appears to break down the liver structure differently than does refrigerated storage.     

USE OF MICROPOROUS PLASTIC FOR IN-PACKAGE CONTROLLED ATMOSPHERES

Ground beef packaged and stored with carbon dioxide-generating reagent packets made from (1) microporous plastic film or (2) non-woven cellulose cloth had significantly (p<0.01) lower bacterial counts than ground meat packaged and stored without packets. There was no significant difference (p>0.05) in bacterial counts of samples stored with the 2 types of packets. Samples with in-package generated carbon dioxide atmospheres had approximately 2–3 days additional shelf-life before spoilage.     

Pittsburgh Labor Market Adjustments In The 1980S: Who Gained and Who Lost?

ABSTRACT: This research describes the distribution of changes in labor market outcomes for individuals in a region disrupted by severe economic restructuring. We use data from western Pennsylvania to compare individuals' 1990 labor market outcomes with those they would have expected had the economic characteristics of the region remained unchanged for the previous 10 years. The comparison between these outcomes provides a measure of the cumulative effect of all labor market changes. We compare the average differences across subgroups of the population to explore how various demographic groups of workers fared. Our analysis suggests that the dynamic forces of economic change have a cumulative effect that differs greatly across subgroups of the population. In particular, older and less educated workers face the largest losses in wages and earnings. As expected, dislocation from a job adds to these economic losses, but the results indicate that age and education have larger negative effects, regardless of dislocation status.