#1 2022

CONTENTS №1

Aristova N.I.  Digital transformation in 2021: achievements and outlook

The paper analyses The level and features of industrial automation business in 2021 is analyzed. Key 2021 innovations underlying the digital transformation of industrial enterprises are overviewed.

Keywords: artificial intelligence, wireless communication, additive technologies, energy-efficient solutions, quantum innovations.

Mindlina O.I.  Predicting product quality indicator

The paper discusses the application of mathematical models, machine learning libraries and big data processing tools (Sklearn, statsmodel, Pandas) for product quality prediction in processing industry.

Keywords: quality prediction, machine learning, physical quantity, measurement, regulatory control systems.

Sergeev A.I., Kornipaev M.A., Rusyaev A.S., Popov A.N.  Investigation of genetic algorithms for parametric synthesis highly automated production systems

The paper examines the problem of parameters fitting for highly automated production systems. Based on computing experiments, it analyzes continuous and discrete genetic algorithms aimed at the search of design parameters optimal subject to equipment load criterion for a flexible production system. The estimate of design parameters is based on computer simulation of production equipment operation. The analysis of computing experiment’s results shows that the highest effectiveness of the global optimum search is attained when a continuous genetic algorithm is used. Based on the estimate obtained, the probability of mutation and the value of the parameter n in the SBX crossover of a continuous genetic algorithm are investigated.

Keywords: continuous and discrete genetic algorithms, parametric synthesis, flexible production system.

Golovkin M.E., Krasnov A.E.  Anisotropy of discrete space in the automation of image analysis process

The paper discusses the anisotropy of the discrete space used in the imagery and image analysis in machine vision. It shows that the rotation of any images specified on the two-dimensional discrete screen results in their significant distortion that hampers their further analysis. In particular, the statistics needed for image description invariantly to their angular orientation are changed. The results of numerical experiments with binary and gray-scale images for assessing the minimal number of etalons needed for invariant description of images for their further recognition are included.

Keywords: discrete space, image analysis, anisotropy, statistics, invariance, histogram, phase portrait.

Drumov I.V., Kamnev M.A., Malyshev G.S., Terekhin G.V.  Neural-network-based defect detection on siliconized graphite products

The results of neural network operation aimed at the classification of defects on siliconized graphite products are analyzed, recommendations on its application are made. The learned network enables partial automation of product inspection procedure with the ensuing inspection time saving that substantiates the study relevance. The paper shows that networks with simple architecture may be used for defect detection. A fully connected network classifier learned for image classification in five categories: crack, cleavage, individual cavity, cavity cluster, and damage-free surface. The architecture of a fully connected network classifier is presented as well as the settings used for network learning.

Keywords: validation set, defects, siliconized graphite, test set, digitization.

Suskin V.V., Filimonov V.V.  Synthesis of computing devices for improving information processing speed

An approach to the synthesis of a custom computing device aimed at information processing speed improvement is presented. The paper discusses the optimization of the structure of multiprocessor computing systems with reference to the effect of the data level on the operation speed, conflict situations while addressing the system’s common resources, as well as the interaction with parallel algorithms, and the design features of custom hardware aimed at maximum computing speed.

Keywords: computing systems’ speed, optimization methods, multiprocessor computing systems, computing speed evaluation, effectiveness, performance improvement, synthesis, parallel computing.

Stetsker B.I.  Edge computing as an integral part of digital transformation

Edge computing concept and architecture are examined. Data operation with edge computing is compared against centralized data processing. Schneider Electric’s solutions for projects with edge computing are discussed.

Keywords: edge computing, micro data center, terminal devices, network infrastructure, information sources, data operation.

Bordyug A.S.  Electrical schematic of a differential semiconductor converter for oxygen and hydrogen control in combustion gas

Due to tighter international regulations and guidelines in sea transportation more ships switch to ecological fuel. To control gas burning, the paper proposes rutile semiconductor converters which enable fuel burning quality control subject to both oxygen content and an overall index dependent on oxygen and hydrogen content. The converter is underlain the relationship between the thermal e.m.f. of nonuniformly heated titanium dioxide and gas composition. An electrical schematic of a differential semiconductor transducer for O₂ and H₂ (hydrocarbons) in combustion products. The paper concludes that the firing of gaseous fuel at low excess oxygen level results in significant improvement of fuel effectiveness.

Keywords: boiler, fuel, firing, semiconductor converter, measurement.

Tikhomirov O.I.  Synchronous electric jet motors for developing energy-efficient IE5 class solutions

Present-day synchronous electric jet motor coupled with a frequency converter is a high energy efficient solution. The paper showcases application advantages of synchronous electric jet motors, offers a solution from ABB, and presents study results of system’s energy efficiency variations dependent on the classes of electric motors applied.

Keywords: energy efficiency, synchronous electric jet motor, frequency converter, payback.

Discussing a Topic…

Monitoring and technical diagnosis in industry

Lapkis A.A., Shvets D.V., Abidova E.A., Dembitsky A.E.  Calculation method for condition inspection of electric drives of shutoff valves

The paper discusses the problem of detecting the defects of driver node and adjusting the moment of electrically driven shutoff valve trip. Such defects cause leaks in seals and failures of critical industrial facilities, such as A-plants. Currently, the challenge is surmounted with the help of special benches, which require dismantling and moving the objects for inspection. To avoid the dismantling of the driver node, the diagnosis based on electric signals is used, but this approach does not ensure the desirable accuracy and the calculation of drive’s torque. The development of a method ensuring the accuracy of torque evaluation is a relevant task. Based on the experimental data, a test pattern for an electric drive is offered as well as an approximating model enabling torque evaluation per indirect measurements if the results of basic tests are available. The new pattern has demonstrated higher approximation accuracy as against the conventional basic test. The results obtained underlie the concept of a portable system combining the basic test bench for electric drives with an instrument for diagnosing electrically driven shutoff valves per electric signals.

Keywords: shutoff valves, diagnosis, torque calculation procedure, driver node’s flaw, electric drive.

Kazymov I.M.  Analysis of power supply system’s state for increased level of power loss

An original method for collecting, processing and analyzing the information about the state of low, medium, and high voltage is proposed. A state estimation algorithm low and medium voltage systems is offered for detecting the increased power loss level. The effectiveness and practical value of the developed method and algorithm are shown; recommendations are made on the application of the results in decision-making system.

Keywords: automated information and measuring system for commercial electricity metering, power supply system’s state analysis, power loss, power supply.

Boyarkov D.A.  An algorithm for risk-oriented control of power network’s technical state

The paper presents an algorithm for risk-oriented control of technical state of power networks. The algorithm prioritizes power network equipment entries in generation programs of power companies as well as specific types of control impacts on electrical equipment. The algorithm comprises of four generic structural blocks: failure probability estimator, power interruption loss calculator, risk assessment and risk management blocks. The algorithm addresses the needs of electric grid companies in reducing the effect of human factor and improving the accuracy of estimates in the development of maintenance and investment programs.

Keywords: power network, failure probability, damage, power supply, risk management, priority, electric grid companies, reliability, monitoring, power user, electrical equipment.

Kereynik Yu.F., Belyaeva V.D.  Modern safety systems for hazardous production facilities

The paper argues that in view of legislative requirements, global trends, and technogeneous incident statistics in Russia, it makes sense to develop industrial safety solutions in two main directions: monitoring systems for industrial facilities and safety management centers (SMC). The functionality of these safety systems is outlined. Special attention is paid to a key function of SMC: the development of digital accident elimination program. Such system would support emergency elimination process and become a primary tool of a manager responsible for emergency elimination.

Keywords: safety systems, hazardous production facility, monitoring of industrial sites, safety management centers, digital accident elimination program.