2687-0517 17 1 2024 1-64

RAR RUS 10-19 Potekhin Vyacheslav slava.potekhin@mail.ru Peter the Great St. Petersburg Polytechnic University Alekseev Anton alekseev.ap@edu.spbstu.ru

St. Petersburg, Russian Federation

Peter the Great St. Petersburg Polytechnic University Kuklin Egor kuklin.ev@edu.spbstu.ru

St. Petersburg, Russian Federation

Belarusian-Russian University Misnik Anton anton@misnik.by

Mogilev, Republic of Belarus

Peter the Great St. Petersburg Polytechnic University Khitrova Yana hitrova.yad@edu.spbstu.ru

St. Petersburg, Russian Federation

Programming of open distributed industrial systems based on the international standard IEC 61499 Today, collaboration in software development and open architectures is changing the fundamental structure of business and reshaping the way organisations operate in a highly competitive environment, forcing them to rethink strategies. Organisations that previously created proprietary systems are beginning to develop open source products to expand the boundaries of the industries in which they operate. In a globalised world, open integrated control systems are becoming increasingly important. Their main goal is to create a balanced, efficient and functional system that integrates various aspects into a coherent whole. The article shows the advantages and disadvantages of using a new approach to programming logic systems based on the international standard IEC 61499 in the field of industrial automation of technological processes. The article analyses basic principles of the IEC 61499 standard, as well as general provisions with the OPAS standard. It also demonstrates a prototype of the control system for the model of the furnace P-101 according to IEC 61499. And provides time characteristics of the software based on IEC 61499 for real-time operating systems. 10.18721/JCSTCS.17101 62-503.55 Industry 4.0 OPAS Cloud DCS Cloud computing Industrial Internet of Things cyber-physical systems https://infocom.spbstu.ru/article/2024.80.1/

RAR RUS 20-32 0000-0001-9587-228X Al Muthanna University College of Engineering Saaudi Ahmed

Iraq

Electronics and Communication Engineering, AL Muthanna University Mansoor Riyadh riyadhdmu@mu.edu.iq

Samawa, Iraq

A study of hyperparameters effect on CNN performance for chest X-ray based COVID-19 detection COVID-19 disease has been spreading around the world for the last four years. Different generations of corona viruses appeared: Alpha-, Beta-, Gamma-, and Delta variants. Thus, COVID-19 changed human lifestyle and affected economic development of many countries. According to clinical studies, most of the positive cases of COVID-19 patients suffer from lung infection. For this, a lot of efforts were aimed at developing fast and accurate detection methods. Thanks to the Deep Learning techniques that facilitate the process of identifying COVID-19 based on the chest images of the patients. X-ray and CT scan images are commonly used to evaluate corona virus lung infection. X-ray images are adopted by many researchers since they place less financial burden on the patient. In this work, we used chest X-ray images to develop eight CNN-based detection models. Three sets of images, i.e., COVID-19, pneumonia and normal cases were used for the training and testing. The performance of each model was optimized based on different hyperparameters to come up with the best results in terms of high detection accuracy, recall, precision and f1 score. These hyperparameters include Number of CNN layers, filters, dense layers, and number of nodes per dense layer. Our findings show that increasing both the CNN layers and number of filters result in high precision and f1 score of the positive samples, while increasing the number of dense layers leads to low precision recall and f1 score. 10.18721/JCSTCS.17102 004 Artificial intelligence COVID-19 CNN Deep learning Chest X-ray https://infocom.spbstu.ru/article/2024.80.2/

SCO RUS 33-43 0000-0002-0950-1333 Peter the Great St. Petersburg Polytechnic University Golovan Olga golovan.olga.andreevna@gmail.com

St. Petersburg, Russian Federation

Analysis of diode mixers using nodal voltage method in generalized matrix form in frequency domain. Part 3: nonlinear distortions A nonlinear distortions analysis method for diode frequency converter circuits is presented. Volterra series method is used to analyze nonlinear distortions. The analysis was carried out for three types of diode frequency converters: balanced, double balanced and triple balanced. Calculation of the 3rd order nonlinear distortion coefficient is presented. The dependences of the 3rd order nonlinear distortion coefficient on the load resistance and on the local oscillator (LO) voltage amplitude were obtained for two LO operation modes: harmonic and pulse. The error between the calculation and simulation results does not exceed 3 dB. It is shown that the dependences of the 3rd order nonlinear distortion coefficient on the load resistance and on the LO voltage amplitude have several maximums and minimums. By varying the values of the load resistance and the LO voltage amplitude it is possible to calculate the minimum achievable value of the 3rd order nonlinear distortion coefficient. 10.18721/JCSTCS.17103 621.3.011.72 diode frequency converters nodal equations method 3-rd order nonlinear distortions Volterra series method balanced mixer double balanced mixer triple balanced mixer https://infocom.spbstu.ru/article/2024.80.3/

RAR RUS 44-53 Peter the Great St.Petersburg Polytechnic University Antonov Alexander antonov@eda-lab.ftk.spbstu.ru

Polytechnicheskaya, 29, St.Petersburg, 195251, Russian Federation

Peter the Great St. Petersburg Polytechnic University Besedin Denis 1310nero@mail.ru

St. Petersburg, Russian Federation

Filippov Aleksey filippov@eda-server.ftk.spbstu.ru Research and comparative analysis of the effectiveness of software and hardware implementations of transposed matrix multiplication The article is devoted to the study and comparative analysis of the software and hardware implementation of the transposed matrix multiplication operation and its modified version, the matrix multiplication transpose. A feature of this study is the use of high-level synthesis tools to obtain and optimize hardware implementations of these operations. The relevance of this study is due to the widespread use of matrix operations, such as transposition and multiplication, to solve various applied problems, the power-law asymptotic complexity of matrix calculations and the lack of data on the effectiveness of using high-level synthesis tools in the tasks of creating hardware devices for matrix calculations. A step-by-step method for synthesizing and optimizing the hardware implementation of these operations is proposed. A comparative study of the software and hardware implementations of these two operations was carried out. It is shown that the gain in performance of hardware implementations is achieved by increasing the degree of parallelism of matrix calculations. Additionally, studies were conducted on the required resources while increasing productivity through parallelization. 10.18721/JCSTCS.17104 004.312.44 hardware implementation performance hardware costs FPGA parallel computing pipelining https://infocom.spbstu.ru/article/2024.80.4/

RAR RUS 54-64 Mozhaisky Military Space Academy Utkin Ivan

St. Petersburg, Russian Federation

Mozhaisky Military Space Academy Shkuropatsky Vitaly vitalius-47@mail.ru

St. Petersburg, Russian Federation

Mozhaisky Military Space Academy Pronikov Alexander

St. Petersburg, Russian Federation

Mozhaisky Military Space Academy Rakov Evgeniy

St. Petersburg, Russian Federation

The study of the vision transformer architecture by explainability methods The article discusses issues of explainability of the operating principles of a machine learning model. As the architecture of the model, one of the types of transformer is considered, the task of which is to classify images based on the popular “ImageNet-1000” dataset. This type of transformer is also called vision transformer and can serve either as a standalone model or as part of a more complex architecture. The explainability methods included activation maps of classes, which were calculated by applying algorithms based on forward and backward propagation of image tensors through the components of the transformer: multi-head attention layers and fully connected multilayer networks. The aim of the work is to increase the explainability of the internal processes of the functioning of the vision transformer by analyzing the obtained activation maps and calculating a metric to evaluate their explainability. The results of the study reveal patterns that reflect the mechanisms of operation of the vision transformer in solving the image classification problem, as well as evaluating the importance of the identified classification features through the use of the explainability metric. 10.18721/JCSTCS.17105 681.3.05 machine learning model explainability visual transformer encoder attention mechanism class activation maps back propagation activation maps https://infocom.spbstu.ru/article/2024.80.5/