Search

In this paper, the problem of order selection for nonlinear dynamical Takagi-Sugeno (TS) fuzzy models is adressed. It is solved by reformulating the TS model in its Linear Parameter Varying (LPV) form and applying an extension of a recently proposed Regularized Least Squares Support Vector Machine (R-LSSVM) technique for LPV models. For that, a nonparametric formulation of the TS identification problem is proposed which uses data-dependent basis functions. By doing so, the partition of unity of the TS model is preserved and the scheduling dependencies of the model are obtained in a nonparametric manner. For the local order selection, a regularization approach is used which forces the coefficient functions of insignificant values of the lagged input and output towards zero.

A stable estimation of the THz model parameters for low SNR configurations is essential to achieve acquisition times required for applications in, e.g., quality control. The deep optimization prior approach was introduced with application to the estimation of material-related model parameters from THz data, which is acquired by a FMCW THz scanning system. Conceptually, this approach estimates the desired THz model parameters by optimizing for the weights of a 3D spatially coupled deep neural network. This approach was verified numerically on various THz parameter estimation problems for synthetic and real data. In this paper, we propose to combine the deep optimization prior approach to the modern 2D blind deblurring method for the FMCW THz image resolution enhancement. The experimental results show that this approach improves the lateral resolution enhancement robustly under low SNR noise condition in comparison to the per-pixel curve fitting method.

The influence of a SARS-CoV-2 infection on inflammatory bowel disease (IBD) has not yet been well characterized and it is unclear whether this requires an adaptation of the immunosuppressive therapy.A national register was established for the retrospective documentation of clinical parameters and changes in immunosuppressive therapy in SARS-CoV-2 infected IBD patients.In total, only 3 of 185 IBD patients (1.6 %) were tested for SARS-CoV-2 infection because of abdominal symptoms. In the course of COVID-19 disease, 43.5 % developed diarrhea, abdominal pain or hematochezia (risk of hospitalization with vs. without abdominal symptoms: 20.0 % vs. 10.6 %, p 0.01). With active IBD at the time of SARS-CoV-2 detection, there was an increased risk of hospitalization (remission 11.2 %, active IBD 23.3 % p 0.05). IBD-specific therapy remained unchanged in 115 patients (71.4 %); the most common change was an interruption of systemic therapy (16.2 %).New abdominal symptoms often appeared in SARS-CoV-2 infected IBD patients. However, these only rarely led to SARS-CoV-2 testing. A high IBD activity at the time of SARS-CoV-2 detection was associated with an increased risk of hospitalization.Der Einfluss einer SARS-CoV-2-Infektion auf chronisch entzündliche Darmerkrankungen (CED) ist bislang nicht gut charakterisiert, und es ist unklar, ob diese eine Anpassung der immunsuppressiven Therapie erfordert.Für die retrospektive Dokumentation klinischer Parameter und Veränderungen einer immunsuppressiven Therapie von mit SARS-CoV-2 infizierten CED-Patienten wurde ein nationales Melderegister etabliert.Insgesamt wurden nur 3 von 185 CED-Patienten (1,6 %) wegen abdomineller Symptome auf eine SARS-CoV-2-Infektion getestet. Im COVID-19-Krankheitsverlauf entwickelten 43,5 % Durchfall, abdominelle Schmerzen oder Hämatochezie (Hospitalisierungsrisiko mit vs. ohne abdominelle Symptome: 20,0 % vs. 10,6 %, p 0,01). Bei aktiver CED zum Zeitpunkt des SARS-CoV-2-Nachweises bestand ein erhöhtes Hospitalisierungsrisiko (Remission 11,2 %, aktive CED 23,3 % p 0,05). Die CED-spezifische Therapie blieb bei 115 Patienten (71,4 %) unverändert; die häufigste Änderung bestand in einer Unterbrechung der systemischen Therapie (16,2 %).Bei mit SARS-CoV-2 infizierten CED-Patienten traten häufig neue abdominelle Symptome bei Infektion auf. Diese führten aber nur selten zur SARS-CoV-2-Testung. Eine hohe CED-Aktivität zum Zeitpunkt des SARS-CoV-2-Nachweises war mit einem erhöhten Hospitalisierungsrisiko assoziiert.

Applications of multistatic sensor arrays, which operate in the terahertz-region, are increasingly gaining importance for non-destructive testing purposes. Continuous enhancements in processing capabilities of CPUs and GPUs allow to generate volumetric images with superior speeds compared to alternative raster-scan systems. However, this requires efficient reconstruction- and correction algorithms. In addition to the system design, the choice of signal processing methods has a decisive influence on the resolution, image quality and measurement duration. In this contribution, we discuss some aspects of this signal processing along with specific examples.

The sterility of drug products intended for parenteral administration is a critical quality attribute (CQA) because it serves to ensure patient safety, and is thus a key requirement by health authorities. Whilst sterility testing is a probabilistic test, the assurance of sterility is a holistic concept including adequate design of manufacturing facilities, process performance, and product design. Container closure integrity testing (CCIT) is necessary to confirm the integrity of a container closure system (CCS), until the end of a products’s shelf life. The new and revised United States Pharmacopeia (USP) informational chapter is a comprehensive guidance on CCI. Nevertheless, practical considerations including the choice of CCIT methods, the acceptance criteria, or the positive control samples (artificial leaks) must be addressed by the pharmaceutical manufacturer. This study is the first to provide a systematic comparison of four commonly used physical CCIT (pCCIT) methods [Helium (He) leak, vacuum decay, laser-based headspace analysis (HSA), and dye ingress] and four commonly used modes of creating artificial leaks (laser-drilled micro holes, copper wire introduced leaks, and two types of capillary leaks). The results from these experiments provide comprehensive data to allow a direct comparison of the capabilities of the individual methods. The results confirmed that the He leak detection method, which is considered the “gold-standard” for pCCIT regarding method sensitivity, indeed demonstrates the highest detection sensitivity (lowest detection limit). In comparison to the dye ingress method, HSA and vacuum decay, also demonstrated better detection sensitivity in our study. Capillary leaks with orifice diameter (capillary leak with flow according to an ideal orifice) and micro holes yielded similar leak rates, whereas capillaries with nominal diameters yielded significantly lower leak rates. In conclusion, method sensitivity cannot be compared by means of a leak diameter, but requires the consideration of multiple impacting factors (e.g. path length, uniformity).

Zusammenfassung Einleitung Der Einfluss einer SARS-CoV-2-Infektion auf chronisch entzündliche Darmerkrankungen (CED) ist bislang nicht gut charakterisiert, und es ist unklar, ob diese eine Anpassung der immunsuppressiven Therapie erfordert. Methodik Für die retrospektive Dokumentation klinischer Parameter und Veränderungen einer immunsuppressiven Therapie von mit SARS-CoV-2 infizierten CED-Patienten wurde ein nationales Melderegister etabliert. Ergebnisse Insgesamt wurden nur 3 von 185 CED-Patienten (1,6 %) wegen abdomineller Symptome auf eine SARS-CoV-2-Infektion getestet. Im COVID-19-Krankheitsverlauf entwickelten 43,5 % Durchfall, abdominelle Schmerzen oder Hämatochezie (Hospitalisierungsrisiko mit vs. ohne abdominelle Symptome: 20,0 % vs. 10,6 %, p Diskussion Bei mit SARS-CoV-2 infizierten CED-Patienten traten häufig neue abdominelle Symptome bei Infektion auf. Diese führten aber nur selten zur SARS-CoV-2-Testung. Eine hohe CED-Aktivität zum Zeitpunkt des SARS-CoV-2-Nachweises war mit einem erhöhten Hospitalisierungsrisiko assoziiert.

How should one regulate a firm when its investment may cause a negative externality? In this paper we present a model on regulating a firm run by a manager and owned by a shareholder. The regulator can impose a penalty on the manager, the shareholder, or both. Our characterization of optimal regulation in the presence of an agency conflict demonstrates several key results. First, optimal regulation can take two forms. In one regulatory strategy fines are imposed on shareholders and they are lower if the shareholders fire the manager following a negative externality. In the other regulatory strategy fines are imposed only on the manager and no firing takes place. Second, as the agency conflict becomes more severe the “no firing” strategy that imposes a fine on the manager becomes more attractive. And third, we find that regulation costs are lower when the manager and the shareholder are separate entities.

In order to address challenges caused by an increasing need for power transmission, embedded bipolar HVDC links utilizing Modular Multilevel Converter (MMC-HVDC) technology present a preferred solution as several operational and dynamic advantages compared to conventional ac grid enforcement are provided. In this paper, an approach to quickly detect and handle internal as well as dc-side faults in a full-bridge bipolar MMC-HVDC link is presented. Due to non-technical reasons and environmental constraints, the system is comprised of mixed overheadline-cable (OHL-cable) transmission and single point solid grounding with dedicated metallic return (DMR). To validate and evaluate the proposed concept, detailed electromagnetic transient (EMT) simulations are provided. These highlight effects of mixed transmission system design on transient voltage and current stresses as well as on converter-internal quantities.

Many established terahertz imaging modalities are on one side restricted by the tradeoff between resolution and field of view such as in the case of focal plane arrays and on the other side suffer from a limited depth of field such as in the case of quasi-optical terahertz imaging configurations. Furthermore, typical scanning solutions require time-consuming measurement procedures and restrict significant potential industrial deployments of terahertz imaging technology. Imaging with sparse multistatic line arrays in combination with digital beam forming (DBF) techniques enables us to overcome these limitations and offer three-dimensional (3D) terahertz image reconstructions of the object. This contribution addresses the design of such terahertz imaging systems from a general point of view with the focus on the design of sparse line arrays, while considering objects scattering properties. Based on this design concept, the realization of a novel highly sparse 3D terahertz imaging system is discussed. The sparse line array of the system is operating within a frequency range from 75 to 110 GHz and is used in combination with a conveyor in order to generate a synthetic sampling aperture. The system is capable to generate 3D terahertz images with tens of megavoxels at feed motions of up to a few 10 cm/s. Also, a sparse array design in regard to an imaging system operating at 240 GHz with integrated SiGe sensor elements is discussed. In addition, three different DBF algorithms are compared in regard to their computational efficiency.

With the development of advanced cutting tool materials, hard turning has become a beneficial alternative to grinding in finishing of highly stressed mechanical components. However, a poor selection of cutting parameters may cause deterioration of the component’s surface integrity and thus its fatigue life. Up to now, the control of surface integrity parameters is not possible as most of these quantities can not be measured during operation. Hence, suitable models are necessary to predict, e.g., the surface hardness and residual stresses induced by hard turning in-process. In order to achieve this, multiple regression and Takagi-Sugeno modeling are applied and compared in this contribution regarding their capability to predict axial and residual stress depth profiles depending on the cutting parameters and the initial hardness of the workpieces. Data from specimen machined on an industrial lathe and measured by X-ray diffraction are used for the case study.

The accurate detection of transitions between appliance states in electrical signals is the fundamental step that numerous energy conserving applications, such as Non-Intrusive Load Monitoring, rely on. So far, domain experts define rules and patterns to detect changes of appliance states and to extract detailed consumption information of individual appliances subsequently. Such event detectors are specifically designed for certain environments and need to be tediously adapted for new ones, as they require in-depth expert knowledge of the environment. To overcome this limitation, we propose a new unsupervised, multi-environment event detector, called MEED, that is based on a bidirectional recurrent denoising autoencoder. The performance of MEED is evaluated by comparing it to two state-of-the-art algorithms on two publicly available datasets from different environments. The results show that MEED improves the current state of the art and outperforms the reference algorithms on a residential (BLUED) and an office environment (BLOND) dataset while being trained and used fully unsupervised in the heterogeneous environments.

Utilizing composite materials instead of metals allows to reduce weight while maintaining mechanical strength. Fiber-Reinforced Thermoplastics (FRT) have very competitive mechanical properties and additionally offer the capability for mass production, which makes the material particularly attractive in the automotive and aircraft sectors. Typical defects in FRT are dominated by fiber shifts and cracks. We show that 3D THz imaging using a FMCW approach can resolve such defects and is therefore a suitable tool for nondestructive testing applications, specifically for the quality control in FRT mass production.

THz imaging and sensing has demonstrated a wide-ranging application potential. However, the transfer of such basic applicability observations to real-world application scenarios is severely obstructed by fundamental limitations imposed by the comparatively long wavelength of this analytic technique. In this presentation, an overview of recent signal processing developments for the enhancement of the analytic performance of THz imaging and sensing systems is presented. The first part of the presentation introduces advanced signal processing techniques to enhance the spectroscopic investigation capability of THz systems. Experiments are performed at particularly difficult application situations, including inter alia very thin material systems or measurements with strongly absorptive features beyond the signal to noise limitations of spectroscopic instrumentation. Model- based numeric procedures for spectroscopic investigation with pulsed THz systems are derived, which enhance the analytic material data quality by two orders of magnitude in comparison to established numeric procedures. Furthermore, computer vision based blind-deconvolution superresolution approaches are introduced, which allow the unassisted increase of imaging resolution beyond the diffraction limit. Experiments performed with a FMCW- based THz imaging system operating from 514 - 640 GHz demonstrate a resolution increase by a factor of 2.3 beyond the diffraction limit, without requiring any prior knowledge on the point-spread function size or shape of the imaging system, but based on a direct analysis of the imaging data of an unknown target sample.

Terahertz (THz) sensing is a promising imaging technology for a wide variety of different applications. Extracting the interpretable and physically meaningful parameters for such applications, however, requires solving an inverse problem in which a model function determined by these parameters needs to be fitted to the measured data. Since the underlying optimization problem is nonconvex and very costly to solve, we propose learning the prediction of suitable parameters from the measured data directly. More precisely, we develop a model-based autoencoder in which the encoder network predicts suitable parameters and the decoder is fixed to a physically meaningful model function, such that we can train the encoding network in an unsupervised way. We illustrate numerically that the resulting network is more than 140 times faster than classical optimization techniques while making predictions with only slightly higher objective values. Using such predictions as starting points of local optimization techniques allows us to converge to better local minima about twice as fast as optimizing without the network-based initialization.

With the rising number of intermittent renewable energy sources and growing energy demand, transfer capabilities are very close to their limits. Consequently, redispatch events occur more frequently and flexibility of conventional generation and storage devices will become increasingly important. The cooperative multi-area optimization strategy presented here enables transmission system operators (TSOs) to dispatch/redispatch interconnected networks securely, while reducing dispatch/redispatch costs. Schedules for storage devices, conventional- and renewable generation are obtained considering network constraints and ramping rates. An optimal schedule for several control areas is attained, including storage operation to achieve congestion relief. The distributed approach preserves control area responsibilities. All participating control areas attain a schedule close to the global optimum. TSOs implement agreements to share resulting profits. The functionality was shown successfully using stressed 14- and 118-node systems. A cross border dispatch with use of storage devices is realized to maintain a high share of renewable energy source (RES) feed-in, while reducing overall dispatch costs.

We analyze why firms use nonintermediated short-term debt by studying the commercial paper (CP) market. Using a comprehensive database of CP issuers and issuance activity, we show that firms use CP to provide start-up financing for capital investment. Firms� CP issuance is driven by a desire to minimize transaction costs associated with raising capital for new investment. We show that firms with high rollover risk are less likely to enter the CP market, borrow less CP, and borrow more from bank credit lines. Further, CP is often refinanced with long-term bond issuance to reduce rollover risk.

Modeling a technical system requires a lot of effort. Especially, when used for simulation, there are high demands on the models regarding approximation accuracy and computational complexity. As building a model based on physical laws is highly time-consuming, system identification is used increasingly. A key problem in system identification is to select an appropriate model, especially when no a priori information of the considered process is available. This paper treats the application of model selection approaches known from linear regression analysis to the identification of a dynamic boost pressure model for the use in a simulation.

Background Spontaneous reports of acute liver injuries (ALI) in patients taking dronedarone triggered an EMA alert in 2011. This study aimed to assess the risk of ALI for class III antiarrhythmic drugs controlling for the use of other potential ALI-inducing drugs. Methods Between 2010 and 2014, consecutive ALI cases (≥50 years-old) were identified across Germany. ALI was defined as a new increase in at least one of the transaminases ≥3 times the upper limit of normal (ULN) or ≥2 ULN if alkaline phosphatase, with (“definite” case) or without (“biochemical” case) suggestive signs/symptoms of ALI, excluding other liver diseases. Recruited community controls were matched to cases on gender, age and inclusion date. Exposure to antiarrhythmic drugs and co-medication up to 2 years before ALI onset was informed by patients and confirmed by physicians' prescriptions. Adjusted Odds Ratios (aOR) were obtained from conditional multivariable logistic regressions, adjusted for a multivariate disease risk score and co-medication. Results 252 cases and 1081 matched controls were included (59.1% females; mean age: 64 years). Exposure to class III antiarrhythmic drugs was 4.0% in cases and 1.5% in controls, aOR = 3.6 (95% CI: 1.6–8.4). Associations with exposure to dronedarone and amiodarone were respectively 3.1 (95% CI: 0.7–14. 8) and 5.90 (1.7–20.0). Restricting the analysis to definite or severe ALI cases did not change these results. Conclusions Class III antiarrhythmic drugs were associated with ALI, amiodarone displaying the highest risk, and results were robust to case definitions. Continued vigilance is needed for patients taking these drugs.

Traditional energy measurement fails to provide support to consumers to make intelligent decisions to save energy. Non-intrusive load monitoring is one solution that provides disaggregated power consumption profiles. Machine learning approaches rely on public datasets to train parameters for their algorithms, most of which only provide low-frequency appliance-level measurements, thus limiting the available feature space for recognition. In this paper, we propose a low-cost measurement system for high-frequency energy data. Our work utilizes an off-the-shelf power strip with a voltage-sensing circuit, current sensors, and a single-board PC as data aggregator. We develop a new architecture and evaluate the system in real-world environments. The self-contained unit for six monitored outlets can achieve up to 50 kHz for all signals simultaneously. A simple design and off-the-shelf components allow us to keep costs low. Equipping a building with our measurement systems is more feasible compared to expensive existing solutions. We used the outlined system architecture to manufacture 20 measurement systems to collect energy data over several months of more than 50 appliances at different locations, with an aggregated size of 15 TB.

Awareness about the energy consumption of appliances can help to save energy in households. Non-intrusive Load Monitoring (NILM) is a feasible approach to provide consumption feedback at appliance level. In this paper, we evaluate a broad set of features for electrical appliance recognition, extracted from high frequency start-up events. These evaluations were applied on several existing high frequency energy datasets. To examine clean signatures, we ran all experiments on two datasets that are based on isolated appliance events; more realistic results were retrieved from two real household datasets. Our feature set consists of 36 signatures from related work including novel approaches, and from other research fields. The results of this work include a stand-alone feature ranking, promising feature combinations for appliance recognition in general and appliance-wise performances.

In this paper, we propose a cost-effective hardware design for acquiring high-resolution energy data from electricity mains to observe appliance switching transients. By utilizing these distinct appliance switching transients, we are interested to predict all appliances operating in a building and also estimate their power consumption. The proposed design fulfills the data acquisition requirements for effective appliance identification from aggregate load signals. Due to its unique and customizable features, the circuit level electric appliance radar (CLEAR) is capable of operating in multiple environments (residential, commercial, and industrial). CLEAR can simultaneously measure voltage and current data streams from a three-phase system at up to 250 kHz to accurately pinpoint individual appliance switching. The first prototype is installed in an office building to collect office level aggregate data. With this aggregate data, the main challenge lies in detecting and isolating similar appliance switching events from switched-mode power supplies (SMPS) utilized by most office appliances such as laptops, LCDs, and printers.