A Hybrid Model-Based Approach on Prognostics for Railway HVAC

Prognostics and health management (PHM) of systems usually depends on appropriate prior knowledge and sufficient condition monitoring (CM) data on critical components’ degradation process to appropriately estimate the remaining useful life (RUL). A failure of complex or critical systems such as heating, ventilation, and air conditioning (HVAC) systems installed in a passenger train carriage may adversely affect people or the environment. Critical systems must meet restrictive regulations and standards, and this usually results in an early replacement of components. Therefore, the CM datasets lack data on advanced stages of degradation, and this has a significant impact on developing robust diagnostics and prognostics processes; therefore, it is difficult to find PHM implemented in HVAC systems. This paper proposes a methodology for implementing a hybrid model-based approach (HyMA) to overcome the limited representativeness of the training dataset for developing a prognostic model. The proposed methodology is evaluated building an HyMA which fuses information from a physics-based model with a deep learning algorithm to implement a prognostics process for a complex and critical system. The physics-based model of the HVAC system is used to generate run-to-failure data. This model is built and validated using information and data on the real asset; the failures are modelled according to expert knowledge and an experimental test to evaluate the behaviour of the HVAC system while working, with the air filter at different levels of degradation. In addition to using the sensors located in the real system, we model virtual sensors to observe parameters related to system components’ health. The run-to-failure datasets generated are normalized and directly used as inputs to a deep convolutional neural network (CNN) for RUL estimation. The effectiveness of the proposed methodology and approach is evaluated on datasets containing the air filter’s run-to-failure data. The experimental results show remarkable accuracy in the RUL estimation, thereby suggesting the proposed HyMA and methodology offer a promising approach for PHM.

View this article on IEEE Xplore

 

Rotation Representations and Their Conversions

A rigid body motion, which can be decomposed into rotation and translation, is essential for engineers and scientists who deal with moving systems in a space. While translation is as simple as vector addition, rotation is hard to understand because rotations are non-Euclidean, and there are many ways to represent them. Additionally, each representation comes with complex operations, and the conversions between different representations are not unique. Therefore, in this tutorial we review rotation representations which are widely used in industry and academia such as rotation matrices, Euler angles, rotation axis-angles, unit complex numbers, and unit quaternions. In particular, for better understanding we begin with rotations in a two dimensional space and extend them to a three dimensional space. In that context, we learn how to represent rotations in a two dimensional space with rotation angles and unit complex numbers, and extend them respectively to Euler angles and unit quaternions for rotations in a three dimensional space. The definitions and properties of mathematical entities used for representing rotations as well as the conversions between various rotation representations are summarized in tables for the reader’s later convenience.

View this article on IEEE Xplore

 

Tool Wear Monitoring Based on Transfer Learning and Improved Deep Residual Network

Considering the complex structure weight of the existing tool wear state monitoring model based on deep learning, prone to over-fitting and requiring a large amount of training data, a monitoring method based on Transfer Learning and Improved Deep Residual Network is proposed. First, the data is preprocessed, one-dimensional cutting force data are transformed into two-dimensional spectrum by wavelet transform. Then, the Improved Deep Residual Network is built and the residual module structure is optimized. The Dropout layer is introduced and the global average pooling technique is used instead of the fully connected layer. Finally, the Improved Deep Residual Network is used as the pre-training network model and the tool wear state monitoring model combined with the model-based Transfer Learning method is constructed. The results show that the accuracy of the proposed monitoring method is up to 99.74%. The presented network model has the advantages of simple structure, small number of parameters, good robustness and reliability. The ideal classification effect can be achieved with fewer iterations.

View this article on IEEE Xplore

 

An Intelligent IoT Sensing System for Rail Vehicle Running States Based on TinyML

Real-time identification of the running state is one of the key technologies for a smart rail vehicle. However, it is a challenge to accurately real-time sense the complex running states of the rail vehicle on an Internet-of-Things (IoT) edge device. Traditional systems usually upload a large amount of real-time data from the vehicle to the cloud for identification, which is laborious and inefficient. In this paper, an intelligent identification method for rail vehicle running state is proposed based on Tiny Machine Learning (TinyML) technology, and an IoT system is developed with small size and low energy consumption. The system uses a Micro-Electro-Mechanical System (MEMS) sensor to collect acceleration data for machine learning training. A neural network model for recognizing the running state of rail vehicles is built and trained by defining a machine learning running state classification model. The trained recognition model is deployed to the IoT edge device at the vehicle side, and an offset time window method is utilized for real-time state sensing. In addition, the sensing results are uploaded to the IoT server for visualization. The experiments on the subway vehicle showed that the system could identify six complex running states in real-time with over 99% accuracy using only one IoT microcontroller. The model with three axes converges faster than the model with one. The model recognition accuracy remained above 98% and 95%, under different installation positions on the rail vehicle and the zero-drift phenomenon of the MEMS acceleration sensor, respectively. The presented method and system can also be extended to edge-aware applications of equipment such as automobiles and ships.

View this article on IEEE Xplore

 

Design, Modeling, and Analysis of a 3-D Spiral Inductor With Magnetic Thin-Films for PwrSoC/PwrSiP DC-DC Converters

A solution architecture for monolithic system-on-chip (SoC) power conversion is in high demand to enable modern electronics with a reduced footprint and increased functionality. A promising solution is to reduce the microinductor size by using novel magnetically-enhanced 3-D design topologies. This work presents the design, modeling, and analysis of a 3-D spiral inductor with magnetic thin-films for power supply applications in the frequency range of 3–30 MHz. A closed-form analytical expression is derived for the inductance, including both the air- and magnetic-core contributions. To validate the air-core inductance model, we implement a 3-D spiral inductor on PCB. The theoretical calculation of air-core inductance is in good agreement with experimental data. To validate the inductance model of the magnetic-core, a 3-D spiral inductor is modeled with Ansys Maxwell electromagnetic field simulation software. A winding AC resistance model is additionally presented. We perform a design space exploration (DSE) to investigate the significance of the 3-D spiral inductor structure. Two important performance parameters are discussed: dc quality factor (Qdc) and ac quality factor (Qac) . Also, a 3-D spiral inductor structure with magnetic thin-films is characterized in Ansys Maxwell to estimate its potential, and a novel fabrication method is proposed to implement this inductor. The measured relative permeability ( μr ) and the magnetic loss tangent ( tan δ ) of Co-Zr-Ta-B magnetic thin-films, developed in-house, are used to simulate the proposed structure. The promising results of the DSE can be easily extended to improve the performance of other 3-D inductor topologies, such as the solenoid and the toroid. The numerical simulations reveal that the 3-D spiral inductor with magnetic thin-films has the potential to demonstrate a figure-of-merit (FOM) that is significantly higher than traditional inductors.

Published in the IEEE Magnetics Society Section of IEEE Access.

View this article on IEEE Xplore

 

Single-Longitudinal-Mode Thulium-Doped Fiber Laser With Sub-kHz Linewidth Based on a Triple-Coupler Double-Ring Cavity

We propose and demonstrate a stable single-longitudinal mode (SLM) thulium-doped fiber laser (TDFL) with a Fabry- Pérot (F-P) fiber Bragg grating (FBG) filter and a triple-coupler based double-ring cavity (TC-DRC) filter. For the first time, this structure of TC-DRC filter is used to select a single mode from dense longitudinal-modes in a ring cavity TDFL. The design and fabricate methods of TC-DRC filter are revealed and the principle of SLM selection is also analyzed in detail. The experimental results demonstrated the good performance of the proposed filter. The central wavelength of proposed TDFL is 2049.49 nm and its OSNR is higher than 35 dB. There is no obvious wavelength drift during the test, and the power fluctuation is less 0.5 dB. The SLM operation is verified through the self-homodyne method, this laser can be stably maintained in a SLM state after operating for one hour under laboratory condition. In addition, the linewidth is measured less than 10 kHz based on the phase noise demodulation method.

View this article on IEEE Xplore

 

An Intraday Market Design for Colombia’s Energy Transition

The massive promotion of intermittent renewable resources as a decarbonization strategy for economies has led to the need to reconsider current market designs in order to facilitate the integration of this kind of production technology. The intraday market has established itself as an efficient mechanism for these purposes, as has been seen in diverse international experiences, principally Europe. The Colombian electricity market is no stranger to the needs of these reforms, as the National Energy and Mining Planning Office of Colombia (UPME in Spanish) estimates that by 2034 almost 30% of the electricity generation matrix will be comprised of intermittent renewable resources. This paper develops several elements to be considered for the implementation of an intraday market in Colombia, backed with quantitative information, contrary to previous studies that have based their recommendations on qualitative elements. The number of discrete intraday sessions, the effect of changing from a sequential energy and reserve allocation scheme to a co-optimized one, as well as the adjustment mechanism – or a balance mechanism – after the gate closure are, among other aspects, analyzed in this study. The computational simulations executed with real data from two operating months with different characteristics in terms of prices, unavailable assets – a liquidity factor for the new market – and system contingencies support the design elements developed in this study. That is, the proposed design – four sections, co-optimization of energy and reserve, and an adjustment mechanism that is not based only on the use of reserves – tested with real data confirm that this proposal is more convenient than the other kind of design from an operating cost and unavailability management perspective. The results reveal market design elements that must be considered in Colombia and serve as input for other countries – particularly Latin American countries – that are in the process of updating their electricity market designs as part of the current energy transition.

View this article on IEEE Xplore

 

Distributed Current Source Method for Modeling of Magnetic and Eddy-Current Fields in Sensing System Design

This paper presents a distributed current source (DCS) method for modeling the eddy current (EC) fields induced in biological or non-ferrous metallic objects in two-dimensional axisymmetric and three-dimensional Cartesian coordinates. The EC fields induced in the objects, magnetic flux density (MFD) in space, and magnetic flux (MF) of the sensing coils are formulated in state-space representation. The harmonic responses of the eddy current fields and electromotive force (EMF) of the sensing coil are formulated in closed-form solutions. The proposed DCS method is applied to design two eddy current sensing systems. The Bio-Differential Eddy Current (BD-EC) sensor distinguishes biological objects, and the Metal-Coaxial Eddy Current (MC-EC) sensor classifies non-ferrous metallic objects. The simulated EC field and EMF are numerically verified by comparing results with finite element analysis. An example is utilized to illustrate the advantage of the DCS method for calculating the MFD, MF, and EMF contributed from the induced ECD in the objects directly, and the EMF generated from each material. The proposed method, along with a prototype of the BD-EC sensor, has been experimentally evaluated on sweep frequency analysis for detecting meat and bone.

View this article on IEEE Xplore

 

Video Based Mobility Monitoring of Elderly People Using Deep Learning Models

In recent years, the number of older people living alone has increased rapidly. Innovative vision systems to remotely assess people’s mobility can help healthy, active, and happy aging. In the related literature, the mobility assessment of older people is not yet widespread in clinical practice. In addition, the poor availability of data typically forces the analyses to binary classification, e.g. normal/anomalous behavior, instead of processing exhaustive medical protocols. In this paper, real videos of elderly people performing three mobility tests of a clinical protocol are automatically categorized, emulating the complex evaluation process of expert physiotherapists. Videos acquired using low-cost cameras are initially processed to obtain skeletal information. A proper data augmentation technique is then used to enlarge the dataset variability. Thus, significant features are extracted to generate a set of inputs in the form of time series. Four deep neural network architectures with feedback connections, even aided by a preliminary convolutional layer, are proposed to label the input features in discrete classes or to estimate a continuous mobility score as the result of a regression task. The best results are achieved by the proposed Conv-BiLSTM classifier, which achieves the best accuracy, ranging between 88.12% and 90%. Further comparisons with shallow learning classifiers still prove the superiority of the deep Conv-BiLSTM classifier in assessing people’s mobility, since deep networks can evaluate the quality of test executions.

View this article on IEEE Xplore

 

Code Generation Using Machine Learning: A Systematic Review

Recently, machine learning (ML) methods have been used to create powerful language models for a broad range of natural language processing tasks. An important subset of this field is that of generating code of programming languages for automatic software development. This review provides a broad and detailed overview of studies for code generation using ML. We selected 37 publications indexed in arXiv and IEEE Xplore databases that train ML models on programming language data to generate code. The three paradigms of code generation we identified in these studies are description-to-code, code-to-description, and code-to-code. The most popular applications that work in these paradigms were found to be code generation from natural language descriptions, documentation generation, and automatic program repair, respectively. The most frequently used ML models in these studies include recurrent neural networks, transformers, and convolutional neural networks. Other neural network architectures, as well as non-neural techniques, were also observed. In this review, we have summarized the applications, models, datasets, results, limitations, and future work of 37 publications. Additionally, we include discussions on topics general to the literature reviewed. This includes comparing different model types, comparing tokenizers, the volume and quality of data used, and methods for evaluating synthesized code. Furthermore, we provide three suggestions for future work for code generation using ML.

View this article on IEEE Xplore