2022 - Repozytorium publikacji - Politechnika Gdańska

Pokaż publikacje z roku

Publikacje z roku 2022

Pokaż wszystkie

Towards Knowledge Sharing Oriented Adaptive Control
- Guixian Li
- Yufeng Xu
- Haoxi Zhang
- Edward Szczerbicki
2022 Pełny tekst CYBERNETICS AND SYSTEMS
In this paper, we propose a knowledge sharing oriented approach to enable a robot to reuse other robots' knowledge by adapting itself to the inverse dynamics model of the knowledge-sharing robot. The purpose of this work is to remove the heavy fine-tuning procedure required before using a new robot for a task via reusing other robots' knowledge. We use the Neural Knowledge DNA (NK-DNA) to help robots gain empirical knowledge and introduce a Knowledge Adaption Module (KAM) utilizing the deep neural networks (DNN) for knowledge reuse. The initial experiment shows that the target robot can adapt to the inverse dynamic model of the source robot via our KAM and reuse the knowledge shared by the source robot.
Towards Open Research Data in the Economics Discipline
- Piotr Kasprzak
- Magdalena Szuflita-Żurawska
2022 Pełny tekst
Nowadays, Open Research Data, as one of the three pillars of Open Science (along with Open Access and Open Scholarly Communication), is gaining enormous attention from different academic and commercial environments. A wide range of scientific disciplines represent and produce different types of data and at the same time, gather different issues and problems in terms of sharing and dissemination research output. This chapter aims to briefly describe the current state of the scientific data for the business and economics disciplines. It presents the opportunities and concerns in those particular scientific fields where very often conducting research is supported not only by public funds but also by business and industry providers. Specific data types (such as financial data) need a particular approach to maintain them, especially in terms of the licensing, preservation and sharing openly. Their proper management can have a significant impact on the knowledge-based economy.
Towards spectral sensitivity curve for two-photon vision mechanism
- Marcin Marzejon
- Zielińska Agnieszka
- Dorota Stachowiak
- Grzegorz Soboń
- Maciej Wojtkowski
- Komar Katarzyna
2022
Abstract Purpose: The perceived brightness of different visible light sources can be compared with photometric units based on the standardized luminosity curves (300-780nm range). As reported previously (PNAS 111(50), pp. E5445-E5454 (2014)), near-infrared (NIR) radiation can cause isomerization of visual pigments by one- or two-photon absorption. The perceived color of the stimulus is red in the case of one-photon vision (1PV) or corresponds to half of the short-pulsed laser wavelength for the two-photon vision (2PV) mechanism. Since no luminosity curves for NIR range, comparison of such sources with photometric units is impossible. This study aims to provide an initial data for the 2PV spectral sensitivity curve (SSC). Methods: We measured the scotopic visibility thresholds at fovea for 4 healthy dark-adapted subjects (2 males, 2 females; 27-43 y/o) between 750-1075nm. Up to 930nm, the total visibility threshold corresponds only to 1PV mechanism. For the 810-930nm, subjects distinguished 2 thresholds: the loss of half of the wavelength hue (2PV threshold) and total visibility (1PV) threshold. Starting from 950nm, only 2PV mechanism caused the visual sensation. The stimulus was a 0.5 deg diameter flickering circle projected in the retina by scanners. Two tunable pulsed light sources: OPO (750-990nm, Frep=76MHz, τp=204fs) and fiber laser (872-1075nm, Frep=51.5MHz, τp=205±9fs) were employed for psychophysical tests. The study complied with the Declaration of Helsinki and was approved by the Ethics Committee of the Collegium Medicum, NCU. Results: The 1PV thresholds agreed with the previous works (JOSA 37(7), pp. 546-554 (1947)). Within the 810-930nm range, a relatively high standard deviation of the 2PV thresholds was observed, as finding the disappearance of half of the wavelength hue is a relatively difficult task. 2PV thresholds slightly differ for OPO and fiber laser, which is explainable by differences in pulse durations and repetition rates. For ~1040nm, the 2PV threshold values agreed to the previous authors’ results. Conclusions: The 2PV sensitivity increased with wavelength of about 1 order of magnitude (810-1075nm range). Starting from 1010 nm, the shape of the obtained 2PV SSC differs from the wavelength-doubled scotopic luminosity curve for 1PV, corrected by the eye media transmittance. The maximum of 2PV SSC was not found. The results allowed us to obtain the first approximation of the 2PV SCC shape up to 1075nm.
Towards spectroscopic monitoring of photoelectrodes: In-situ Raman photoelectrochemistry of a TiO2/prussian blue photoanode
- Konrad Trzciński
- Mariusz Szkoda
- Zuzanna Zarach
- Mirosław Sawczak
- Andrzej Nowak
2022 ELECTROCHIMICA ACTA
Here, novel in-situ Raman photoelectrochemical measurements are performed. The obtained results have proved that it is possible to track the progress of a photoelectrochemical reaction that takes place on a semiconducting electrode using the spectroscopic method. As an exemplary system, the Ti/TiO2/Prussian blue electrode is investigated. Since TiO2 is an n-type semiconductor, it cannot act as an efficient anode in dark conditions. Thus, once electrochemically reduced Prussian blue deposited on TiO2 is not reoxi- dized during anodic polarization. However, Everitt’s salt oxidation may be possible when photoexcitation of TiO2 with UV LED radiation takes place at a potential which is more anodic than the flat band poten- tial of the semiconductor. This process is successfully monitored in-situ using Raman spectroscopy. The measurements performed at different conditions allow to distinguish between processes induced by po- larization and illumination (or both). The proposed method of photoelectrode characterization through a real-time monitoring approach can be further developed and utilized for other photoelectrochemical systems.
Training of Deep Learning Models Using Synthetic Datasets
- Zdzisław Kowalczuk
- Jan Glinko
2022
In order to solve increasingly complex problems, the complexity of Deep Neural Networks also needs to be constantly increased, and therefore training such networks requires more and more data. Unfortunately, obtaining such massive real world training data to optimize neural networks parameters is a challenging and time-consuming task. To solve this problem, we propose an easy-touse and general approach to training deep learning models for object detection and instance segmentation without being involved in the generation of real world datasets. In principle, we generate and annotate images with open-source software and 3D models that mimic real life objects. This approach allows us significantly reduce the effort required to gather pictures as well as automatize data tagging. It is worth noting that such synthetic datasets can be easily manipulated, e.g. to reduce the texture bias that often occurs in the resulting trained convolutional networks. Using the Mask R-CNN instance segmentation model as an example, we demonstrate that a network trained on the synthetic dataset of kitchen facilities shows remarkable performance on the validation dataset of real-world human-annotated photos. We show that our approach helps to bridge the domain gap between pre-trained models and their specific applications. In summary, such synthetic datasets help overcome the problem of acquiring and tagging thousands of images, while reducing the time and labor costs associated with the preparation of an appropriate real dataset
Transformation of Energy Markets: Description, Modeling of Functioning Mechanisms and Determining Development Trends
- Michał Pietrzak
- Marta Kuc-Czarnecka
2022 Pełny tekst ENERGIES
One of the key contemporary economic and social issues today is the global energy transition. Energy transition processes are having a significant impact on the development of world economies, increasing their TFP and leading to an increase in their level of innovation through the transfer of myriad new technologies. These processes also contribute to an increase in foreign direct investment and, consequently, an increase in the level of business investment, the competitiveness of economies and changes in the labor market. Additionally, contributing to the systematic and dynamic development of the energy transition are the significant increase in the wealth of the population, the change in the degree of social and income inequality, the change in consumption patterns and the significantly increased consumption of energy by households, which until re-cently relied exclusively on the use of energy from conventional sources. It should be noted that the ongoing energy transition processes most strongly affect the development of energy markets, the largest of which are the electricity primary fuel markets. Currently, energy markets represent an increasingly significant aspect of modern economies in terms of business investment, the share of the sector’s output in GDP, as well as research and development. In recent years, the renewable energy sector has also been gaining importance as a natural complement to the two aforementioned markets. All the above-mentioned aspects of the development of modern economies point to the need to take a fresh look at the development and functioning of energy markets. Of particular importance seems to be the analysis of changes in the prices of electricity and primary fuels and the relationship between these markets and the renewable energy mar-ket. Equally important are analyses to identify development trends already occurring in the energy markets and to make predictions about the formation of these trends in the fu-ture. Such identified studies should provide valuable guidance for the purposes of con-ducting current energy policy and creating institutional and legal conditions for the development of energy markets. Conclusions from research on energy markets also provide substantive arguments for the assumptions of global energy strategies, as well as the energy strategy of individual countries.
Transformational Leadership and Acceptance of Mistakes as a Source of Learning: Poland-USA Cross-Country Study
- Wioleta Kucharska
- Maciej Kucharski
- Hafeez Ur Rehman
2022 Pełny tekst
This study explores the influence of transformational leadership on internal innovativeness mediated by mistakes acceptance, including country and industry as factors to be considered and gender and risk-taking attitude as moderators. General findings, primarily based on the US samples (healthcare, construction, and IT industry), confirmed that transformational leadership and internal innovativeness are mediated by mistakes acceptance and strengthened by employees' risk-taking attitude. It is reflected in the Polish healthcare sample. On the other hand, IT and construction industry samples revealed that a low mistakes acceptance level among risk-taking managers under transformational leadership might be their hidden form of change resistance. Regarding gender issue, this study showed that male employees driven by transformational leaders accept mistakes as a source of learning better than women, but the exceptions are healthcare and IT industries in Poland. The developed empirical model is based on a sample composed of 2,160 Polish and American knowledge workers applying OLS regression using SPSS PROCESS macro software.
Transformational leadership for researcher’s innovativeness in the context of tacit knowledge and change adaptability
- Wioleta Kucharska
- Teresa Rebelo
2022 Pełny tekst International Journal of Leadership in Education
This study explores how a learning culture supported by transformational leadership influences tacit knowledge sharing and change adaptability in higher education and how these relations impact this sector’s internal and external innovativeness. The empirical model was tested on a sample of 368 Polish scientific staff using the structural equation modeling (SEM) method. Then results were expanded by applying OLS regression using SPSS PROCESS macro. Findings revealed that tacit knowledge sharing and change adaptability driven by learning culture are vital links connecting transformational leadership with innovativeness. Moreover, change adaptability was revealed to be a critical factor mediating between tacit knowledge sharing and innovativeness (external and internal). Tacit knowledge sharing and transformational leadership are seen as fully mediated by the mistake acceptance factor in learning culture. Besides, this study provided empirical evidence that higher education institutions must adapt to change constantly and evaluate their internal processes to deliver a higher level of innovative work visible externally. It also shows that leaders supporting a smooth flow of tacit knowledge sharing are central to scientific development because they foster adaptability and innovativeness (external and internal). Furthermore, the influence of the mistakes acceptance component on tacit knowledge sharing is revealed to be moderated by gender.
Transformator z tyrystorowym przełącznikiem zaczepów jako obiekt sterowania
- Piotr Sępka
- Maciej Brzycki
- Artur Cichowski
2022 Pełny tekst
Artykuł jest wynikiem realizacji studenckich projektów, których obiektem badań były transformatory z tyrystorowym podobciążeniowym przełącznikiem zaczepów (PPZ), znajdujące się w Laboratorium LINTE^2. Pierwszy projekt obejmował analizę stanów przejściowych zachodzących podczas przełączania tyrystorów. W referacie przedstawiono opracowany model tyrystorowego przełącznika zaczepów wraz z przykładowymi wynikami. Drugi projekt skutkował utworzeniem translatora komunikacyjnego w oparciu o sterownik WAGO PFC200, który umożliwia reprezentację modelu danych urządzeń niezgodnych z normą IEC 61850, jako zgodne z tym standardem
Transmisja w kanale hydroakustycznym w warunkach NLOS
- Agnieszka Czapiewska
- Ryszard Studański
- Andrzej Żak
2022 Pełny tekst Przegląd Elektrotechniczny
Transmisja danych w kanale hydroakustycznym w warunkach NLOS (ang. Non-Line-Of-Sight) jest szczególnie problematyczna. Dlatego podjęto próbę opracowania bezprzewodowej komunikacji podwodnej i jej przebadanie w warunkach NLOS. Zastosowano modulację MFSK (ang. Multiple Frequency-Shift Keying), przesyłając pojedynczy bit na dwóch nośnych, oraz odbiór zbiorczy. Uzyskane wyniki potwierdziły przydatność opracowanej techniki do transmisji danych w kanale hydroakustycznym w warunkach NLOS.
Trees as a solar control measure for southern-oriented street frontages. Analysis of a selected street model for a humid continental climate
- Katarzyna Zielonko-Jung
- Justyna Janiak
2022 Pełny tekst URBAN ECOSYSTEMS
The present study is aimed at the analysis of possibilities for shading southern frontage of street oriented along the E-W axis by the single row of trees, parallel to the southern elevations. The effectiveness of solar control shading was tested depending on the geometric relationships between trees and buildings. Numerical simulation analyses were conducted in Rhinoceros® program for the street located in humid continental climate in city Płock, Poland (52°32′50 “N 19°42’00 “E), for the day of the highest degree of total solar radiation in the year i.e. June 7th, during hours: 8.00a.m - 5.00 p.m. The analysis has proved that a row of 20–25 year old Sycamore Maple ‘Rotterdam’ in the street 30 m wide and 18 m high (H/W = 0.6), can provide solar protection for the southern frontage, especially when trees are located no more than 4 m away. Location of greenery within the range of 4 to 5 m from the buildings leads to a radical reduction in the possibility of shading the wall surfaces (at 5 m to 0%). Over 90% of the shading area of the ground floor façade walls was found when trees were within the distance 2 and 3 m away from the building.
Trends in the new generation of green solvents in extraction processes
- Patrycja Janicka
- Justyna Płotka-Wasylka
- Natalia Jatkowska
- Aneta Chabowska
- Michel Y. Fares
- Vasil Andruch
- Massoud Kaykhaii
- Jacek Gębicki
2022 Pełny tekst Current Opinion in Green and Sustainable Chemistry
Analytical chemistry, like other scientific fields, has undergone a number of changes to make it more consistent with the concept of sustainable development. Among the various steps of chemical analysis, without a doubt, sample preparation is the bottleneck in regard to following a green protocol, especially in terms of solvent consumption. Therefore, many attempts have been made to improve the environmental friendliness of this stage, mainly through the developing approaches for miniaturized sample preparation as well as application of new green solvents. This review offers a brief discussion of current trends in analytical applications that have been less studied and discussed: a new generation of green solvents, such as bio-based solvents, supercritical fluids, and liquefied. We believe that this mini review is a good starting place for readers interested in the future of green analytical chemistry.
Trendy w kształtowaniu nowoczesnych dzielnic badawczo-rozwojowych i przemysłowo-logistycznych
- Karolina Krośnicka
2022
Tekst opisuje aktualne trendy w kształtowaniu przestrzennym nowoczesnych dzielnic badawczo-rozwojowych oraz przemysłowych i logistycznych.
Tribological Properties of Thermoplastic Materials Formed by 3D Printing by FDM Process
- Jacek Łubiński
2022 Pełny tekst
The dataset entitled 3D printed ABS thermoplastic vs. steel. Dry sliding wear test in constant load & velocity ring on flat configuration. Test parameters: print layer thickness and orientation. Test symbol: 019_h_4 contains: the time base (expressed in seconds and minutes), the friction torque for sliding friction, rotational velocity of the counter – specimen (velocity of sliding), friction coefficient, load in the friction contact interface, specimen temperature. Tests were conducted in an in-house developed tribological tester, the Tribometr PT-1. The test belongs to a testing program on the influence of technological parameters in 3D printing from thermoplastic materials by FDM process on friction in sliding with a steel counter – specimen without added lubrication.
Tripping of F-type RCDs for sinusoidal residual current with superimposed smooth DC component
- Stanisław Czapp
- Hanan Tariq
2022
Recent trends in green energy development make that photovoltaics and electric vehicles are applied on an increasing scale. In both photovoltaic and electric vehicle charging installations, a significant value of DC component of the earth fault current may appear, which is a challenge, in particular, for commonly used residual current devices (RCDs). This paper presents results of the laboratory test on the operation of F-type RCDs – a type that is just beginning to appear in low-voltage installations. It has been checked how the tripping threshold of RCDs changes, if a smooth DC component is superimposed on the sinusoidal alternating current. It has been found that the F-type RCDs can operate very well even in the presence of a very high value of DC component.
Trójfazowy 4-gałęziowy falownik SiC w napędzie z wysokoobrotowym silnikiem indukcyjnym
- Marek Adamowicz
- Sebastian Giziewski
- Zbigniew Krzemiński
2022
Zastosowanie tranzystorów SiC w falownikach napędów wysokoobrotowych powoduje nowe wyzwania, szczególnie odnośnie generowanego w falowniku napięcia common mode (CM) o wysokiej częstotliwości. Wysokoczęstotliwościowe napięcie CM wchodzi w interakcję z pojemnościami pasożytniczymi falownika, kabla i silnika wymuszając przepływ prądów CM, które są źródłem strat w filtrach pasywnych oraz powodują zakłócenia pomiarów prądów, nieakceptowalne z punktu widzenia działania przekształtnika. Ze wzrostem częstotliwości rośnie poziom emisji zaburzeń elektromagnetycznych spowodowany rezonowaniem obwodu CM, a tym samym niekorzystnie rośnie masa, gabaryt i koszt dławików CM. W referacie zaproponowano zastosowanie 4-gałęziowego falownika SiC z metodą sterowania zapewniającą eliminację napięcia CM celem poprawy właściwości napędu wysokoobrotowego.
Trust and Distrust in e-Democracy
- David Duenas Cid
- Tomasz Janowski
- Robert Krimmer
2022
In the digital government research literature, the concept of trust is typically used as a precondition for the adoption of digital technology in the public sector or an outcome of a roadmap leading up to such adoption. The concept plays a central role in many decisions linked to the planning, adoption and management of the public sector technology. In contrast, the concept of distrust is almost neglected in such literature but, when approached, it appears as the opposite logical side of the trust-distrust dichotomy. However, we conjecture that the path to building trust is different than the path to building distrust and both concepts should be regarded as different theoretical constructs. The workshop aims to prove this conjecture drawing on the insights from the field of e-democracy including internet voting. Given its technical breadth, political and societal implications, and layers of complexity, e-democracy is a good benchmark for exploring this topic. The workshop will confront the main conjecture using examples, research and experiences contributed by participants, conceptual and methodological tools introduced by the organizers, and a discussion shared by everybody. The outcome – elements that help build trust or distrust in various forms of e-democracy – will have practical and theoretical implications, aiming at further research by the participants and collective publication of the results in a special issue in a top ranked journal, preferably Government Information Quarterly.
Trust in Electronic Word-of-Mouth (eWOM): the concept and the most important determinants
- Anna Drapińska
2022
eWOM and trust in it is now the most important factor in determining buyers' purchasing behavior. Familiarity with the determinants of trust establishes the formation of appropriate marketing activities. However, the growing number of new forms of eWOM and its increasingly complex and multidimensional nature have created a lot of confusion. The article aims to organize the concepts related to trust in eWOM, to indicate the most important known determinants and future research directions. Thanks to the analysis of the literature, it was shown that future research on trust in eWOM should take into account more advanced models that include the multidimensionality of eWOM and trust, as well as individual product categories and customer characteristics.
Trustworthy Applications of ML Algorithms in Medicine - Discussion and Preliminary Results for a Problem of Small Vessels Disease Diagnosis.
- Maria Ferlin
- Zuzanna Klawikowska
- Julia Niemierko
- Małgorzata Grzywińska
- Arkadiusz Kwasigroch
- Edyta Szurowska
- Michał Grochowski
2022 Pełny tekst
ML algorithms are very effective tools for medical data analyzing, especially at image recognition. Although they cannot be considered as a stand-alone diagnostic tool, because it is a black-box, it can certainly be a medical support that minimize negative effect of human-factors. In high-risk domains, not only the correct diagnosis is important, but also the reasoning behind it. Therefore, it is important to focus on trustworthiness which is a concept that includes fairness, data security, ethics, privacy, and the ability to explain model decisions, either post-hoc or during the development. One of the interesting examples of a medical applications is automatic SVD diagnostics. A complete diagnosis of this disease requires a fusion of results for different lesions. This paper presents preliminary results related to the automatic recognition of SVD, more specifically the detection of CMB and WMH. The results achieved are presented in the context of trustworthy AI-based systems.
t-SNE Highlights Phylogenetic and Temporal Patterns of SARS-CoV-2 Spike and Nucleocapsid Protein Evolution
- Gaik Tamazian
- Andrey Komissarov
- Dmitry Kobak
- Dmitry Polyakov
- Evegeny Andronov
- Sergei Nechaev
- Sergey Kryzhevich
- Yuri Porozov
- Eugene Stepanov
2022
We propose applying t-distributed stochastic neighbor embedding to protein sequences of SARS-CoV-2 to construct, visualize and study the evolutionary space of the coronavirus. The basic idea is to explore the COVID-19 evolution space by using modern manifold learning techniques applied to evolutionary distances between variants. Evolutionary distances have been calculated based on the structures of the nucleocapsid and spike proteins.