#4 2022

CONTENTS №4

Wagner O.V.  Digital steelmaker’s assistant based on computer vision and machine learning technologies

The paper outlines the development and testing phases for a custom solution for assisting steelmakers in slagging and steel output. The digital assistant is based on computer vision and machine learning technologies. The paper discusses the challenges faced by the developers and the ways to overcome them. It provides a block diagram of the assistant’s computing platform. The development outlook of Datana’s solutions is included.

Keywords: digital steelmaker’s assistant, computer vision, machine learning, IR camera, slagging, steel output, visualization.

Tkachenko A.L., Mescheryakov V.A., Denisova L.A.  Design of an analytic support system for the educational process in a technical university

The paper discusses the development of an analytic system for a technical university. The system improves the effectiveness of the educational process in the field of industrial automation. A method for automatic classification of text documents in the university’s electronic workflow system is developed on the basis of machine learning techniques and a digital document model. A decision-making support software for the employees of the graduates’ professional growth center is offered. The software helps to select enterprises for practical training of students and their further employment.

Keywords: information and educational environment, workflow automation, decision-making support, personnel training for industrial enterprises.

Fyodorov A.A.  A method for eddy current loss control in a stator core

The assembly of stator cores from electric steel sheets should avoid short circuits resulting in higher losses. The paper examines a quality control method for stator core production. Theoretical and practical parts of the testing by means of circular magnetization techniques are described, test data for the core of a 2000 V motor are presented.

Keywords: stator core, loss control, heat monitoring, specific loss in a core, eddy currents, magnetic loss, electric steel.

Discussing a Topic…

Automation of oil, gas, and petrochemical industries

Nikulina I.V.  Industrial robotics – Made in Russia

The paper overviews the most recent domestic designs of industrial robots: manipulators, palletizers, delta robots, collaborative robots, and mobile robot systems. Their performance data, functionalities, and possible applications are outlined.

Keywords: industrial robots, manipulators, delta robots, collaborative robots, and mobile robot systems.

Koltygin D.S., Sedel’nikov I.A., Ulyanov A.D.  Developing a procedure for manipulator envelope modeling

The paper offers a procedure for developing working envelopes of various manipulating robots with serial structure as well as the envelope visualization software. The space where the robots move should be specified to avoid overlapping with ensuing process upsets, equipment damage, and health hazards. For building the working envelope for a manipulator with the serial structure, an algorithm based on direct and kinematics problem is developed. It ensures the determination of a pixel array in three coordinate planes with their subsequent unification in the manipulator’s working envelope. The envelope curve is further determined by means of the link-by-link assembly of an equations set with subsequent search of allowed values. A numerical experiment was undertaken for checking solution correctness. The result was compared with design parameters of the same space and validated with lab equipment.

Keywords: robotics, direct and inverse kinematics problems, manipulator envelope.

Aristova N.I., Chadeev V.M.  Updating the project of a flexible robot cell for manufacturing cartridge type components

The paper discusses the development of a flexible robot cell for cartridge type component machining. At the project implementation phase, it became necessary to replace a robot for these operations at minimum average cost and in compliance with the spec. A replacement robot had to be selected and application efficiency substantiated. To that end, digital simulation techniques were used, which enabled the analysis of automation efficiency for the specified set of production operations.

Keywords: flexible robot system, industrial robot, digital model, analysis of production operations efficiency.

Tsuprikov A.A.  Intelligent adaptive control system for oil-well drilling

The paper presents a control system for oil-well drilling. The system enables the optimal control of machine drilling, hydraulic flushing, and round-trip operations. It recognizes the geological material in the bottomhole and adapts to it thus ensuring environmental protection drilling safety owing to early detection and prevention of more than 20 types of emergencies and accidents including hole blows and gushers. The system employs: (i) a sinking velocity model with optimal adjustment of three parameters subject to three criteria, (ii) a wash liquor flowrate model allowing for the drilling chips volume, and (iii) adaptive control strategy.

Keywords: intelligent system, adaptive optimal control, drilling model, trouble-free drilling.

Tikhomirov L.I., Romanov D.S.  Digitization of mining enterprise: key objectives and first steps

The paper presents a present-day approach to the improvement of process and operations efficiency at a mining enterprise. The approach is based on the development of unified centralized data warehouses and the implementation of digital systems, which ensure production performance indices. A case study of an ongoing digitization project at a mining enterprise is included.

Keywords: digitization, centralized data warehouse, mining enterprise.

Vermel’ V.D., Yevdokimov Yu.Yu., Kacharava I.N., Nikulenko A.A., Rozin N.V., Smotrov A.V., Tellin I.S., Trifonov I.V., Usov A.V., Chernyshev L.L.  Automation of the correction of natural frequency responses of blades of the models of helicopter’s lifting rotors for resonance elimination in experimental research

For a standard structure of a large-scale helicopter rotor’s blade model made of polymer composite materials, a parameter-oriented finite-element model is developed for calculating key dynamic characteristics (natural frequencies and waveforms) in the ABAQUS industrial software suite. Based on the calculation results, the resonant revolution diagram is developed, and nominal design points for weight fixing and the masses of balance weights are determined for frequency response detuning to eliminate the resonance.

Keywords: large-scale aerodynamic model, helicopter rotor’s blade dynamic behavior, rotation speed range, resonance, resonant revolutions diagram.

Buynevich M.V., Izrailov K.E., Pokusov V.V., Romanov N.E.  A method for variability classification of vulnerabilities in program code. Part 2. Machine learning-based automation

The problem of variative classification of vulnerabilities in program code is solved with the help of machine learning. The first part of the article discussed the first four steps of the procedure. The second part focuses on the remaining for steps, namely: the description of vulnerabilities, their textual evidences, identification and classification. For implementing the most difficult last step, the relevant scholarly articles were overviewed. The approaches underlying the automated classification technique were pointed out, their merits and drawbacks were emphasized. A procedure based on machine learning with respect to the classification of certain vulnerability characteristics based on their textual description is developed. A prototype implementing Word2Vec and TF-IDF techniques as well as five classifiers is developed. The prototype was tested against vulnerability descriptions from the Federal Service for Technical and Export Control (FSTEC of Russia). The most effective combinations of techniques and classifiers were detected.

Keywords: software, program code vulnerabilities, variability classification, machine learning, prototype, experiment.

Zakharov N.A.  Additive technologies in industry

Additive technologies are classified, examples of their application in aircraft industry and healthcare are cited. The marketplace of 3D printing services and consumables is outlined.

Keywords: additive manufacture, 3D printing, selective laser melting, micro stereolithography, marketplace.

Shchapova L.V.  Automated measuring of arterial pressure

An algorithm which calculates systolic and diastolic pressure for automated monitoring of cardiovascular system status is developed on the basis of the support vector machine. The Gaussian radial basis functions were selected as the machine’s kernel functions. The model’s parameter vector includes the minimum and maximum values of photoplethysmography signal, the minimum of the second derivative of the photoplethysmography signal, and the amplitude of the R wave of ECG signal. The model was validated over 155 palpitation intervals with the average absolute error 2,47 ± 1,75 mm Hg for the systolic pressure and 3,04 ± 2,41 for the diastolic one. For 156 degrees of freedom, the t-value was 1,01×10^-7 for the systolic pressure and 3,87×10^-5 for the diastolic one.

Keywords: support vector machine, arterial pressure, photoplethysmography, electrocardiography (ECG).