Technical University of Sofia
Recent publications
The aim of this study is to compare effective dose (E) estimations based on different methods for patients with recurrent computed tomography (CT) examinations. Seventeen methods were used to determine the E of each phase as well as the total E of the CT examination. These included three groups of estimations: based on the use of published E, calculated from typical or patient-specific values of volume computed tomography dose index (CTDIvol) and dose-length product (DLP) multiplied by conversion coefficients, and based on patient-specific calculations with use of software. The E from a single phase of the examination varied with a ratio from 1.3 to 6.8 for small size patients, from 1.2 to 6.5 for normal size patients, and from 1.7 up to 18.1 for large size patients, depending on the calculation method used. The cumulative effective dose (CED) ratio per patient for the different size groups varied as follows: from 1.4 to 2.5 (small), from 1.7 to 4.3 (normal), and from 2.2 up to 6.3 (large). The minimum CED across patients varied from 38 up to 200 mSv, while the variation of maximum CED was from 122 up to 538 mSv. Although E is recommended for population estimations, it is sometimes needed and used for individual patients in clinical practice. Its value is highly dependent on the method applied. Individual estimations of E can vary up to 18.1 times and CED estimations can differ up to 6 times. The related large uncertainties should always be taken into account.
This paper has been devoted to an analysis of geometry from an engineering point of view. Has been based 23 statements about geometry contribution to physics, mathematics, engineering, as well as to other sciences. Geometry propagation in all types of objects and processes (such as material, energetic and information/signal ones), as well as in everything and everywhere (all-encompassing geometry) has been shown. Special attention is paid to geometry importance, emphasizing its crucial role at each stage of any technical object design (which predetermines geometry as the basis of design), as well as in other cases. It is proposed to establish a strict line of demarcation between the science "General Geometry" as "Geometry of Everything" with a construct (any object considered as a geometric structure) on the one hand, and with a subject (a study of spatial structures and construct relations with an assigned aim) – on the other hand, and to develop them separately.
Let [ · ] be the fioor function. In this paper, we show that when 1 < c < 37/36, then every sufficiently large positive integer N can be represented in the form where p 1 , p 2 , p 3 are primes close to squares.
In this paper, the parameters that shape online reputation of a company and the two-way relation between the effect reputation has on consumer behavior and the effect consumer behavior has on reputation are examined. The contribution of this work lies in the fact that it concerns Greek consumers. Moreover, it is important to highlight the particularities of Greek consumers and to confirm or disprove international trends in their case. At the same time, issues related to the extraction and classification of elements related to reputation are highlighted. The answers given through this research and the questions that arise contribute to the formulation of effective strategies for the management of online reputation by the companies, even in cases of attacks and threats.
An emerging mode of refrigerated preservation has recently been advertised as an allegedly game-changing means to improve the overall sustainability of global frozen food industry. Isochoric low-temperature pressure-aided supercooling (named ‘isochoric freezing’) creates exciting expectations, associated with reduced or eliminated ice crystallisation and resulting finer food structure, along with savings of cold energy for phase change. A simplified technology assessment has been carried out by comparing this emerging modality with the conventional isobaric freezing for a typical lab-scale scenario. Except several niche applications, in general the lauded method appears to underperform classical isobaric freezing in terms of applicability, and overall usage of resources and energy.
The pyrolysis of wheat straw in order to produce biochar for soil amendment is a potential strategy for producing environmental friendly fertilizers capable of boosting soil fertility, increasing carbon storage, and lowering greenhouse gas emissions. However, straw biochar’s potential to influence these aspects may vary depending on its properties. Our study sought to investigate biochar from wheat straw from three different regions in Bulgaria. A specially designed set up was used for the biochar production. Three pyrolytic temperatures (300, 400, and 500 °C) were applied, resulting in nine biochar samples. The specific characteristics included moisture content, volatile substances content, ash content, fixed carbon content, and joint ash and carbon content, and they were determined for each sample. The chemical content, resulting in 17 chemical elements and compounds, was measured and analysed. The results obtained showed that the produced straw biochar has the potential to be used as a fertilizer and soil supplement.
This manuscript proposes a new unified approach for the analysis of voltage source inverters that summarizes and unifies the operating modes of a whole class of inverters powered by a voltage source: resonant, aperiodic and voltage source inverters. The study was performed on a full-bridge circuit of a series RLC inverter with reverse diodes, operating in aperiodic mode. Based on the community of processes in the power circuits, the expressions for the current through the load and the voltage of the capacitor in condensed form with their initial phases are determined. These basic ratios are presented in a normalized form according to the control frequency, thus summarizing the operation of the inverters with a control frequency below and above the quasi-resonant. A methodology for designing a voltage source inverter was developed, considered as a special case of a series RLC inverter with inverse diodes when operating in aperiodic mode and super-quasi-resonant frequency. The reliability of the obtained results was verified by comparison with the use of the classical methodology for design and computer simulations. The presented approach is useful from a methodological point of view, as it allows us, with a unified approach and through general mathematical expressions, to describe and study the processes in a significant part of DC/AC converters.
Let [ · ] be the floor function. In this paper we show that every sufficiently large positive integer N can be represented in the form N = [p_1log p_1] + [p_2 log p_2] + [p_3 log p_3] where p_1, p_2, p_3 are prime numbers. We also establish an asymptotic formula for the number of such representations, when p_1, p_2, p_3 do not exceed given sufficiently large positive number.
In this paper, a flexibility adequacy assessment of the South East Europe region countries is being presented. Novel technology integration is being considered in order to provide more flexibility resources to the power system to absorb more renewable energy. A flexibility analysis based on the International Energy Agency methodology provided an overall estimation of the flexibility needs and resources of the Bulgarian and Cypriot power systems. Additionally, several flexibility indices have been calculated providing indications of the potential that both systems have to serve more volatile renewable energy sources without jeopardizing the balancing requirements for frequency regulation and security of supply. A detailed algorithm has been developed, in close cooperation with the national stakeholders in Bulgaria and Cyprus, in order to simulate the variations in demand and generation for the following years and calculate statistical indices for flexibility, such as the Insufficient Ramping Rate expectation and Flexibility Residual, apart from the traditional Loss of Load Expectation used in adequacy studies.
The electron-impact widths for 13 Si II multiplets, belonging to the 3s3p(3Po)nl configuration, have been calculated by using the modified semiempirical method. Since a number of these lines is in visible, they are of interest and for analysis and synthesis of stellar Si II lines, abundance determination and modelling of stellar atmospheres.The obtained results have been compared with available experimental and theoretical data and used for the investigation of regularities within supermultiplets.
In recent years, the IoT has emerged as the most promising technology in the key evolution of industry 4.0/industry 5.0, smart home automation (SHA), smart cities, energy savings and many other areas of wireless communication. There is a massively growing number of static and mobile IoT devices with a diversified range of speed and bandwidth, along with a growing demand for high data rates, which makes the network denser and more complicated. In this context, the next-generation communication technology, i.e., sixth generation (6G), is trying to build up the base to meet the imperative need of future network deployment. This article adopts the vision for 6G IoT systems and proposes an IoT-based real-time location monitoring system using Bluetooth Low Energy (BLE) for underground communication applications. An application-based analysis of industrial positioning systems is also presented.
Future train control systems are expected to support high speed railways, to be automated and intelligent. European Railway Traffic Management System (ERTSM) and European Train Control Systems (ETCS) provide a standard aimed at optimization of driving, constant control on the effect of actions taken for train safety and activating of train braking in case of emergency. Radio Block Center (RBC) is a central control in the ETCS ground topology and is responsible for the security of each train in its area. Currently, the RBC is proprietary equipment with monolithic design. In this paper, we propose a new disaggregated open architecture of RBC which enables embedding intelligence. The proposed architecture of intelligent RBC is based on separation of nonreal time functions and near real time functions. A use case illustrates the approach applicability. Some implementation issues are discussed.
The Future Railway Mobile Communication System (FRMCS) has emerged as a worldwide standard for railway communication. This technology enables the operational efficiency and safety of railways to be improved by providing mission critical communications, machine-type communication for the railway system on board, in addition to trackside telemetry and broadband connectivity for passengers. Different equipment types, users, and functional identities can be involved in communication, and each of them is uniquely identified. Identity management is an important part of the security functions provided by the FRMCS system. This paper presents a service-oriented approach to identity management functionality, enabling service composition for railway applications and service virtualization. This paper studies functionality for the initial registration and subsequent deregistration of railway devices, users, and their functional identities, in addition to the transfer of the registered identities between different FRMCS serving areas while the train moves. Two FRMCS services that follow the principles of representational state transfer architecture are proposed. Services’ functionality is illustrated by use cases, data types, and application programming interfaces that enable services to be interacted with. Identity registration status models are developed, formally described, and mathematically verified. Discussion of the applicability of the proposed services for the implementation of FRMCS security and safety functions is provided. The presented service-oriented approach features a satisfactory level of flexibility and versatility.
Nowadays, air pollution is an important problem with negative impacts on human health and on the environment. The air pollution forecast can provide important information to all affected sides, and allows appropriate measures to be taken. In order to address the problems of filling in the missing values in the time series used for air pollution forecasts, the automation of the allocation of optimal subset of input variables, the dependency of the air quality at a particular location on the conditions of the surrounding environment, as well as automation of the model’s optimization, this paper proposes a deep spatiotemporal model based on a 2D convolutional neural network and a long short-term memory network for predicting air pollution. The model utilizes the automatic selection of input variables and the optimization of hyperparameters by a genetic algorithm. A hybrid strategy for missing value imputation is used based on a combination of linear interpolation and a strategy of using the average between the previous value and the average value for the same time in other years. In order to determine the best architecture of the spatiotemporal model, the architecture hyperparameters are optimized by a genetic algorithm with a modified crossover operator for solutions with variable lengths. Additionally, the trained models are included in various ensembles in order to further improve the prediction performance—these include ensembles of models with the same architecture comprising the best architecture obtained by the evolutionary optimization, and ensembles of diverse models comprising the k best models of the evolutionary optimization. The experimental results for the Beijing Multi-Site Air-Quality Data Set show that the proposed spatiotemporal model for air pollution forecasting provides good and consistent prediction results. The comparison of the suggested model with other deep NN models shows satisfactory results, with the best performance according to MAE, based on the experimental results for the station at Wanliu (16.753 ± 0.384). Most of the model architectures obtained by the optimization of the model hyperparameters using the genetic algorithm have one convolutional layer with a small number of kernels and a small kernel size; the convolutional layers are followed by a max-pooling layer, and one or two LSTM layers are utilized with dropout regularization applied to the LSTM layer using small values of p (0.1, 0.2 and 0.3). The utilization of ensembles from the k best trained models further improves the prediction results and surpasses other deep learning models, according to MAE and RMSE metrics. The used hybrid strategy for missing value imputation enhances the results, especially for data with clear seasonality, and produces better MAE compared to the strategy using average values for the same hour of the same day and month in other years. The experimental results also reveal that random searching is a simple and effective strategy for selecting the input variables. Furthermore, the inclusion of spatial information in the model’s input data, based on the local neighborhood data, significantly improves the predictive results obtained with the model. The results obtained demonstrate the benefits of including spatial information from as many surrounding stations as possible, as well as using as much historical information as possible.
Fifteen 4-methyl-1,2,3-thiadiazole-based hydrazone derivatives 3a-d and sulfonyl hydrazones 5a-k were synthesized. They were characterized by 1H-NMR, 13C NMR, and HRMS. Mycobacterium tuberculosis strain H37Rv was used to assess their antimycobacterial activity. All compounds demonstrated significant minimum inhibitory concentrations (MIC) from 0.07 to 0.32 µM, comparable to those of isoniazid. The cytotoxicity was evaluated using the standard MTT-dye reduction test against human embryonic kidney cells HEK-293T and mouse fibroblast cell line CCL-1. 4-Hydroxy-3-methoxyphenyl substituted 1,2,3-thiadiazole-based hydrazone derivative 3d demonstrated the highest antimycobacterial activity (MIC = 0.0730 µM) and minimal associated cytotoxicity against two normal cell lines (selectivity index SI = 3516, HEK-293, and SI = 2979, CCL-1). The next in order were sulfonyl hydrazones 5g and 5k with MIC 0.0763 and 0.0716 µM, respectively, which demonstrated comparable minimal cytotoxicity. All compounds were subjected to ADME/Tox computational predictions, which showed that all compounds corresponded to Lipinski's Ro5, and none were at risk of toxicity. The suitable scores of molecular docking performed on two crystallographic structures of enoyl-ACP reductase (InhA) provide promising insight into possible interaction with the InhA receptor. The 4-methyl-1,2,3-thiadiazole-based hydrazone derivatives and sulfonyl hydrazones proved to be new classes of lead compounds having the potential of novel candidate antituberculosis drugs.
In this work, we present a new hierarchical decomposition aimed at the decorrelation of a cubical tensor of size 2^n, based on the 3D Frequency-Ordered Hierarchical KLT (3D FO-HKLT). The decomposition is executed in three consecutive stages. In the first stage, after adaptive directional vectorization (ADV) of the input tensor, the vectors are processed through one-dimensional FOAdaptive HKLT (FO-AHKLT), and after folding, the first intermediate tensor is calculated. In the second stage, on the vectors obtained after ADV of the first intermediate tensor, FO-AHKLT is applied, and after folding, the second intermediate tensor is calculated. In the third stage, on the vectors obtained from the second intermediate tensor, ADV is applied, followed by FO-AHKLT, and the output tensor is obtained. The orientation of the vectors, calculated from each tensor, could be horizontal, vertical or lateral. The best orientation is chosen through analysis of their covariance matrix, based on its symmetry properties. The kernel of FO-AHKLT is the optimal decorrelating KLT with a matrix of size 2x2. To achieve higher decorrelation of the decomposition components, the direction of the vectors obtained after unfolding of the input tensor in each of the three consecutive stages, is chosen adaptively. The achieved lower computational complexity of FO-AHKLT is compared with that of the Hierarchical Tucker and Tensor Train decompositions.
Institution pages aggregate content on ResearchGate related to an institution. The members listed on this page have self-identified as being affiliated with this institution. Publications listed on this page were identified by our algorithms as relating to this institution. This page was not created or approved by the institution. If you represent an institution and have questions about these pages or wish to report inaccurate content, you can contact us here.
1,580 members
Malinka Ivanova
  • Faculty of Applied Mathematics and Informatics
Vassil Galabov
  • Department of Industrial Automation
Mariya Aleksandrova
  • Department of Microelectronics
Zahari Zarkov
  • Department of Electrical Machines
8, Kliment Ohridski Blvd., BG-1000, Sofia, Bulgaria
Head of institution
Prof. Ivan Kralov, PhD, DSc