Estimation of material functions using system identification techniques

Material properties of an elastic material are characterized by the elastic modulus, which is real-valued and constant. For viscoelastic materials, such as plastics and polymers, the relationship between stress and strain is instead dynamic, and characterized by the complex-valued and frequency-dependent complex modulus. It is in this paper described how system identification techniques can be used to determine the complex modulus using strain data from wave propagation experiments on a test specimen. Modeling, derivation of estimators, and analysis of their numerical and statistical properties are included. Identifiability and experimental design are examined in some detail. Several practical examples are presented using real-world data, and a number of extensions are outlined.     

PBCDD and PBCDF from incineration of waste-containing brominated flame retardants

Brominated organic flame retardants comprise a large, heterogeneous group of compounds that are useful but also potentially damaging to the environment. In this investigation, three common brominated flame retardants were co-combusted with an artificial municipal solid waste in a pilot-scale fluidized bed incinerator. Combustion conditions ensured degradation of the flame retardants, but stable brominated organic compounds such as polybrominated dibenzo-p-dioxins and polybrominated dibenzofurans were formed in the cooling flue gases. The incineration of fuels that contain both chlorine and bromine leads to the formation of a complex mixture of polybrominated/chlorinated dibenzo-p-dioxins and dibenzofurans. More bromination than chlorination was observed in these experiments, and the chlorination levels increased when bromine was added to the fuel. The most favored reactions led to the formation of dibromodichlorodibenzo-p-dioxin and dibromodichlorodibenzofuran. Theoretical calculations show that there is more Br2 than Cl2 in the flue gas when the two halogens are present at equimolar levels, because chlorine is mainly found in the form of HCl. This may explain the higher bromination level. BrCl is also present, which drives the chlorination observed when bromine is added.     

On a method for model structure selection in system identification

Bohlin has recently suggested an appealing method for model structure selection. It is shown that the method can be interpreted as a modified F-test. The minimal loss in a larger model structure is substituted by an estimate. The estimate is obtained by minimization of a quadratic approximation of the loss function.     

Passive air sampling using semipermeable membrane devices at different wind-speeds in situ calibrated by performance reference compounds

Semipermeable membrane devices (SPMDs) are passive samplers used to measure the vapor phase of organic pollutants in air. This study tested whether extremely high wind-speeds during a 21-day sampling increased the sampling rates of polycyclic aromatic hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs), and whether the release of performance reference compounds (PRCs) was related to the uptakes at different wind-speeds. Five samplers were deployed in an indoor, unheated, and dark wind tunnel with different wind-speeds at each site (6-50 m s(-1)). In addition, one sampler was deployed outside the wind tunnel and one outside the building. To test whether a sampler, designed to reduce the wind-speeds, decreased the uptake and release rates, each sampler in the wind tunnel included two SPMDs positioned inside a protective device and one unprotected SPMD outside the device. The highest amounts of PAHs and PCBs were found in the SPMDs exposed to the assumed highest wind-speeds. Thus, the SPMD sampling rates increased with increasing wind-speeds, indicating that the uptake was largely controlled by the boundary layer at the membrane-air interface. The coefficient of variance (introduced by the 21-day sampling and the chemical analysis) for the air concentrations of three PAHs and three PCBs, calculated using the PRC data, was 28-46%. Thus, the PRCs had a high ability to predict site effects of wind and assess the actual sampling situation. Comparison between protected and unprotected SPMDs showed that the sampler design reduced the wind-speed inside the devices and thereby the uptake and release rates.     

Candida glabrata prosthesis infection following pyelonephritis and septicaemia

A 48-year-old woman with no cardiovascular risk factors was admitted to the hospital because of acute dyspnea. At 27-year-old, she developed Hodgkin's disease, that was successfully treated with splenectomy, combined chemotherapy (nitrogen mustard, vincristine, procarbazine, prednisone-MOPP regimen) and radiotherapy (4500 rads). At 43-year-old the lymphoma relapsed and she had further chemotherapy with doxorubicin, bleomycin, vinblastina and dacarbazine. After this treatment, she had an episode of pulmonary edema, attributed to doxorubicin acute cardiotoxicity. She responded to digitalis and diuretics and was discharged with an electrocardiogram (ECG) showing left bundle branch block and a normal echocardiogram. The patient enjoyed good health for several years and 4 months before the present admission the ECG and echocardiogram were unchanged. On this admission there were signs of left ventricular failure with acute pulmonary edema, and a new soft apical murmur (3-4 Levine). The patient required endotracheal intubation and high doses of diuretics, digitalis and vasodilators. The cardiac enzymes were negative, the serial ECGs confirmed left bundle branch block, while the echocardiogram showed moderate to severe mitral regurgitation, akinesia of the interventricular septum and inferior wall with dilation of the left ventricle. A previous silent myocardial infarction was suspected. After recovery, she underwent cardiac catheterization confirming akinesia of the interventricular septum and inferior wall with moderate mitral regurgitation, while coronary angiography showed a critical ostial stenosis of the right coronary artery. In view of a dipyridamole-thallium scan negative for myocardial viability, reperfusion was not attempted. With changes in radiotherapeutic techniques, the incidence of radiation-induced heart disease (pericarditis, myocarditis, conduction abnormalities and, rarely, occlusive coronary artery disease) is declining. Nevertheless, after irradiation of the chest and mediastinum a longterm cardiological follow-up is useful in selecting patients at higher risk of radiation-induced coronary artery disease, who will eventually require coronary angiography and reperfusion intervention.     

Asymptotic bias of the high-order autoregressive estimates of sinusoidal frequencies

Fitting a high-order autoregressive model to data is often used to estimate the frequencies of sinusoids from their noisy measurements. This paper analyzes the asymptotic bias of the frequency estimates obtained by this method. It is shown that the bias decreases significantly with increasing model order.The work of B. Friedlander was supported by the Office of Naval Research under contract No. N000014-84-C-0408.     

Separate versus Simultaneous Saccharification and Fermentation of Two‐Step Steam Pretreated Softwood for Ethanol Production

Abstract

In two previous studies, optimal conditions were identified for two‐step steam pretreatment of SO₂‐ and H₂SO₄‐impregnated softwood. In the present study the yield of sugar and ethanol was determined in a process development unit where pretreatment was performed in a 10‐L reactor and simultaneous saccharification and fermentation (SSF) or enzymatic hydrolysis (EH) were performed in 30‐L reactors. The study showed that a steam pretreatment reactor should be larger than 2 L to yield acceptable results. Two pretreatment combinations were studied. In the H₂SO₄ case, the first pretreatment step was at 180°C for 10 min with 0.5% H₂SO₄ and the second step at 210°C for 2 min with 1% H₂SO₄. In the SO₂ case, first step was at 190°C for 2 min followed by a second step at 210°C for 5 min. The concentration of SO₂ was 3% in both steps. EH and SSF were performed on the whole slurry after the second pretreatment step to determine the yield of sugars and ethanol. The liquid after the first pretreatment step was also analyzed and fermented. When SSF and EH were performed at the same dry matter content and enzymatic activity, the ethanol yield in SSF exceeded the yield obtained with EH in both pretreatment cases, even when 100% yield in the fermentation step was assumed. Thus SSF is a better process if yield is the main priority. Comparison of the yields with the two acid catalysts showed higher yields with SO₂ in both SSF and EH. The overall ethanol yield following SSF of SO₂‐impregnated and pretreated wood reached 81% of the theoretical, that is, 357 liters per metric ton of dry raw material.     

CD94/NKG2 is the predominant inhibitory receptor involved in recognition of HLA-G by decidual and peripheral blood NK cells

Prior studies demonstrated that NK cells isolated from adult peripheral blood kill the HLA-A-, HLA-B-, and HLA-C-deficient B lymphoblastoid cell line 721.221, but many are unable to kill 721.221 cells transfected with HLA-G, a molecule expressed preferentially on fetal cytotrophoblasts. To determine the biologic relevance of this recognition, we established NK cell clones from the placenta and demonstrate that these NK cells also were unable to kill 721.221 cells expressing HLA-G. Recognition of HLA-G by NK cells was prevented in the presence of anti-CD94 mAb, implicating CD94/NKG2 as the predominant inhibitory NK cell receptor for HLA-G used by decidual NK cells. In contrast, mAbs against the killer cell inhibitory receptors recognizing HLA-Cw3-related, HLA-Cw4-related, or HLA-Bw4 ligands did not affect NK cell killing of the HLA-G transfectants.     

Decentralized array processing using the MODE algorithm

Centralized methods for source location using sensor arrays have computational and communication burdens that increase significantly with the number of sensors in the array. Therefore, these methods may not be usable in the applications involving very large arrays. In such applications, the data processing may need to be decentralized. This paper introduces two methods for decentralized array processing, based on the recently proposed MODE algorithm. For prescribed nonoverlapping subarrays, both methods are shown to be statistically optimal in the sense that asymptotically they provide the most accurate decentralized estimates of source location parameters. The problem of subarray selection to further optimize the estimation accuracy is only briefly addressed. The two methods are intended for different types of applications: the first should be preferred when there exist significant possibilities for local processing or for parallel computation in the central processor; otherwise the second method should be preferred. The accuracy of the two decentralized methods is compared to the centralized Cramér-Rao bound, both analytically and numerically, in order to provide indications about the loss of accuracy associated with decentralized processing.The work of P. Stoica was supported by a grant from the Royal Swedish Academy of Engineering Sciences and by the Swedish Research Council for Engineering Sciences under contract 91-676. The work of A. Nehorai was supported by the Air Force Office of Scientific Research under grant no. AFOSR-90-0164, and by the Office of Naval Research under grant no. N00014-91-J-1298.on leave at the Department of Applied Electronics, Chalmens University of Technology, 2-41296 Göthensberg, Sweeden.     

Some properties of the output error method

The output error identification method is studied in various respects. The stationary points of the associated loss function are investigated. Sufficient conditions for a unique local minimum are given. The loss functions can be minimized using a quasilinearization algorithm. Such an algorithm will give good local convergence. It is, however, shown that global convergence does not always occur. The output error method is also compared with some other estimation methods from the accuracy point of view. It is proved that a prediction error method will give better accuracy. An instrumental variable technique may give better or worse accuracy depending on the actual noise correlation.