We present a novel framework for multivariate time series representation learning based on the transformer encoder architecture. The includes an unsupervised pre-training scheme, which can offer substantial performance benefits over fully supervised downstream tasks, both with but even without leveraging additional unlabeled data, i.e., by reusing existing data samples. Evaluating our several public datasets from various domains and diverse characteristics, we demonstrate that it performs...

Medium-voltage (MV) SiC devices have been developed recently which can be used for three-phase MV grid tie applications. Two such devices, 15 kV insulated-gate bipolar transistor (IGBT) and 10 MOSFET, opened up the possibilities of looking into different converter topologies distribution interface. These in drives, active filter applications, or as front end solid-state transformers (SSTs). The transformerless intelligent power substation (TIPS) is one application these devices. TIPS...

Abstract Background Artificial intelligence (AI) and large language models (LLMs) can play a critical role in emergency room operations by augmenting decision-making about patient admission. However, there are no studies for LLMs using real-world data scenarios, comparison to being informed traditional supervised machine learning (ML) models. We evaluated the performance of GPT-4 predicting admissions from department (ED) visits. compared ML both naively when few-shot examples and/or...

Magnetic iron oxide nanoparticles and near-infrared (NIR) fluorescent single-walled carbon nanotubes (SWNT) form heterostructured complexes that can be utilized as multimodal bioimaging agents. Fe catalyst-grown SWNT were individually dispersed in aqueous solution via encapsulation by oligonucleotides with the sequence d(GT)15, enriched using a 0.5 T magnetic array. The resulting nanotube show distinct NIR fluorescence, Raman scattering, visible/NIR absorbance features, corresponding to...

Existing temporal pattern mining assumes that events do not have any duration. However, in many real world applications durations, and the relationships among these are often complex. These modeled using a hierarchical representation extends Allen's interval algebra. this is lossy as exact cannot be fully recovered. In paper, we augment with additional information to achieve lossless representation. An efficient algorithm called IEMiner designed discover frequent patterns from interval-based...

The solid-state transformer (SST) is a promising power electronics solution that provides voltage regulation, reactive compensation, dc-sourced renewable integration, and communication capabilities, in addition to the traditional step-up/step-down functionality of transformer. It gaining widespread attention for medium-voltage (MV) grid interfacing enable increases energy penetration, and, commercially, SST interest traction applications due its light weight as result medium-frequency...

In the past decade, we have witnessed explosive growth in number of low-power embedded and Internet-connected devices, reinforcing new paradigm, Internet Things (IoT). The low power wide area network (LPWAN), due to its long-range, low-cost communication capability, is actively considered by academia industry as future wireless standard for IoT. However, despite increasing popularity `mobile IoT', little known about suitability LPWAN those mobile IoT applications which nodes varying degrees...

A dual active bridge (DAB) is a zero-voltage switching (ZVS) high-power isolated dc-dc converter. The development of 15-kV SiC insulated-gate bipolar transistor device has enabled noncascaded medium voltage (MV) DAB It offers simple control compared to cascaded topology. However, compact-size high frequency (HF) transformer significant parasitic capacitances for such voltage. Under and dV/dT switching, the parasitics cause electromagnetic interference loss. They also pose additional...

Blockchain and big data are the emerging technologies that on highest agendas of firms. These significantly expected to transform ways in which business as well firm run. verge increasing expectation distributed ledgers, would keep firms away from struggling challenges. The concept has been used up by various other concepts help secure interpret information. ideal solutions offered these shall address challenges management for analytics. In addition that, provides its own consensus method,...

Locating the stator interturn (SIT) fault on motor winding structure adds an important feature in diagnosis. This motivates to study effects of SIT location induction machine. In this paper, a simple yet accurate stationary reference frame q-d-0 model SIT-faulted machine, including parameter, is developed. The fundamental components functions (WFs) are used calculate machine inductances for proposed model. These variables rotor-position-independent expressions and severity location. as...

Cascaded converters are generally used for medium-voltage (MV) grid-connected applications due to the limitation in voltage rating of available silicon (Si) power devices. These find application active filters, STATCOM or as front end solid state transformers at distribution levels. The high wide bandgap semiconductor devices have enabled grid connected operation noncascaded converters. This results density, less number switching devices, and efficiency three-phase MV interface. also control...

Medium-voltage (MV) silicon carbide (SiC) devices such as the 15-kV SiC N-insulated gate bipolar transistor (IGBT) have better thermal withstanding capability compared with (Si)-based devices. These also lower switching and conduction losses at high frequencies power levels, respectively. The maximum safe operating junction temperature for IGBT is 175 °C. This enables density design of MV converters using this device. Heat sink forced air cooling considered dissipating heat generated during...

A large number of time series forecasting models including traditional statistical models, machine learning and more recently deep have been proposed in the literature. However, choosing right model along with good parameter values that performs well on a given data is still challenging. Automatically providing set to users for dataset saves both effort from using trial-and-error approaches wide variety available optimization. We present AutoAI Time Series Forecasting (AutoAI-TS) provides...

The search for efficient storage and transportation systems of natural gas has led to the exploration hydrate-based technologies as a potential solution. hydrate formation development have propensity occurrence under suitable conditions. enhancement in growth kinetics through usage surfactants was attributed enhanced heat mass transfer, reactor design, ability promote crystals but limited result generation foam during process dissociation. Apart from surfactants, studies emphasized...

This paper introduces a scalable Anomaly Detection Service with generalizable API tailored for industrial time-series data, designed to assist Site Reliability Engineers (SREs) in managing cloud infrastructure. The service enables efficient anomaly detection complex data streams, supporting proactive identification and resolution of issues. Furthermore, it presents an innovative approach modeling infrastructure by utilizing Large Language Models (LLMs) understand key components, their...

Distributed generators (DGs) are being increasingly connected at medium voltage level distribution networks. It is likely that different types of DGs, i.e. wind turbine (WTGs), photovoltaic solar systems, etc, may be on the same feeder. Most modern DGs equipped with power electronic converters their terminals, which act as sources harmonic injection into network. Any feeder system will have its own network resonant frequencies. In this paper, impact WTG connections modes impedance and total...

Modern motor protection relays are capable of long-term failure monitoring and diagnostics based on information collected by a powerful data acquisition system. This includes predicting lifetime motors. paper presents an advanced algorithm performing such tasks. It is applicable to different types rotating electric machines depending their design. The components as follows: stator turn fault detection rotor bar detection. Detection methods calculating the cross-coupled impedance analyzing...

The time integral of the output voltage vector a three-phase inverter is often termed flux vector. This paper addresses control four-leg sine-wave having an LC filter at its output, by controlling in three dimensions. Flux has property that resonance actively damped loop alone. Furthermore, switching action inherently regulates rapidly against dc-bus variations. inverters finds several applications four-wire systems. presents modulation method for feeding unbalanced and nonlinear loads. All...

This paper summarizes the different steps that have been undertaken to design medium voltage power converters using state-of-the-art 15 kV SiC N-IGBTs. The 11 switching characterization results, high dv/dt gate driver validation, and heat-run test results of IGBT at 10 kV, 550 W/cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> (active area) recently reported as individual topics. In this paper, it is attempted link all these topics...

We apply a network approach to identify genes associated in<italic>cis</italic>or in<italic>trans</italic>with copy-number alterations in breast cancer pathogenesis.

High voltage high frequency (HF) transformer provides the isolation between and low dc link voltages in dual active bridge (DAB) converters. Such DAB converters are finding wide applications as an intermediate DC-DC converter transformerless intelligent power substation (TIPS), which is proposed alternative for conventional distribution-transformer connecting 13.8 kV 480 V grids. The design of HF used stage such application very challenging considering required isolation, size cost. In this...