#5 2022

CONTENTS №5

Discussing a Topic…

On the development of discrete digital manufacturing

Martinova L.I., Fokin N.N., Belyakov P.A.  Digital twin based development and debugging of NC system for cross-platform milling cycle of rag wheel's teeth

The paper presents the development and implementation of a milling cycle for rag wheel's teeth. The cycle was developed and debugged on the basis of a digital twin of AxiOMA Control multifunctional CNC system using the programming syntax of a higher-level language. The cycle's parameters are analyzed and its variables are formed. The cycle algorithm is developed and exemplified in a software module using two macros. The first one is responsible for the component part's geometry, whereas the other one controls the machining process. The problem of machining cycle reuse in various NC systems is examined, and a solution is proposed. Based on the identified variables and the algorithm, a teeth milling cycle was developed using the digital twin of AxiOMA Control.
Keywords: control program, NC system, digital twin, machining cycles, macroprogramming, cycle variables

Kovalyov I.A., Nenarokomov M.D., Chervonnova N.Yu., Kvashnin D.Yu.  Combined method for identification of NC machine operator

The paper examines the methods for process operator identification by the example of local binary pattern algorithm and artificial neuron networks techniques. The approaches are compared with each other. A unified software module is developed and tested on a laboratory bench.
Keywords: process data acquisition, Siamese neural networks, NC system, operator identification, local binary pattern algorithm.

Martinov G.M., Zakharov A.S.  The features of historical data implementation in OPС UA server for NC system

The paper discusses the structure of data acquisition and storage mechanism using OPC UA protocol in NC machining process. A software implementation of the OPC UA server for AxiOMA Control CNC system using open62541 open source library was undertaken. The solution was tested using historical data and UaExpert full-featured client.
Keywords: CNC system, OPC UA, client/server, historical data.

Korablyov A.V., Turkov V.A.  Matrix aircraft assembly production in Industry 4.0 paradigm

Matrix production concept is presented. The application of digital twin technology in the design and control of matrix production is shown by the example of aircraft industry.
Keywords: robotization, digital twins, modeling, off-line programming, matrix production.

Kurnosenko A.E., Vlasov A.I., Dyomin A.A.  Discrete simulation of hands-on operations of electronics package assembly on circuit board

Within the framework of a digital twin concept for assembling, the paper offers a digital model for hands-on operations of electronics package assembly. It presents the results of PCB assembler's workbench model investigation using anthropomorphic dummies. The variants of workbench organization aimed at better accessibility of functional assemblies were examined. The results of hands-on operations model embedment into the general model of electronics assembly line are shown. They include the assessment of equipment and personnel load dependent on the number workbenches employed for hands-on operations.
Keywords: electronics assembly, simulation modeling, digital twin, anthropomorphic dummy.

Tyurin O.G., Tyurin G.O., Kuz'menko D.V., Omelyanchuk S.V.  Process control system for group pressing of pyrotechnic elements from high-sensitivity phosphorous-containing pellets

The present-day production of pyrotechnic elements includes explosive pressing operation. To eliminate the hazard for employees, the paper proposes to use an unmanned automated pressing line with a conveying system based on industrial manipulator. Its design was modified to meet production requirements. The system uses dust-ignition-proof hardware, including the batch-weighing feeder.
Keywords: group pressing, robotization, batch weighing, explosion protection.

Ivanovsky A.N., Cherniy S.G.  Implementation of intelligent identification components for the draught survey of cargo ships

Draught survey is a procedure for assessing cargo weight at marine and river vessels based on their draughts. It is especially relevant for bulk carriers, where it is the only alternative to weighing. An automated draught survey control system proposed in the article allows to minimize the influence of the human factor on the measurement results, increase the accuracy and reduce the time of cargo mass calculation. A key feature of the system is the way to determine the draught using neural networks and a linear filter as well as an unmanned aerial vehicle for automated draught measurements.
Keywords: draught survey, cargo operations, ship's draught, sagging and hogging, automatization.

Polyantseva K.A.  High load platform for aggregation and analysis of unstructured roadbed condition data

The paper presents a high load platform for aggregation and analysis of unstructured roadbed condition data. The platform warns road services and their staff about the need in timely roadbed maintenance. A neural network algorithm for roadbed flaw categorization is developed. Based on the instrument readings coming from terminals, the algorithm categorizes the flaws into four classes: cracks, ruts, breaches, and others. The classification algorithm is based on Mask R-CNN neural network architecture. Batch-normalization layers were added for decreasing the overall number of layers and improving the classification accuracy (up to 92.01%). Microserver architecture was chosen for meeting serviceability, simplicity and modularity requirements. Each micro resource is deployed in its own Docker container.
Keywords: high loaded applications, flaw detection, neural networks, unstructured data, roadbed.

Avdeeva Z.K., Grebenyuk E.A., Kovriga S.V.  Developing medium-term monthly forecasts of feedstock prices based on expert and quantitative information

The development of a yearly feedstock price forecast with monthly breakdown is discussed. Such forecasts are a necessary element of industrial enterprise planning activity. However, the forecast uncertainty caused by external factors grows with the planning horizon extension. To reduce the uncertainty, the paper offers a hybrid model for monthly development and update of the annual price forecast. Along with the indicators predicted and the prices affecting the time series forecast, the proposed model allows for the information obtained from expert judgement processing by means of the cognitive map of the situation. The predictive model uses the ensembles of multi-dimensional time series models, which allow for the intensity of the influence of the key factors estimated on the basis of expert information. For correcting the predicted index at the prediction horizon an algorithm was developed based on the signals of situational and digital monitoring. The process of forecast update at the prediction horizon is illustrated by the example of ferrous scrap price prediction for 2019 based on situational and digital monitoring signals.
Keywords: medium-term forecast, situation monitoring, fuzzy cognitive maps, analysis of price time series.

Ivakin Y.A., Semenova E.G., Shashurin A.E., Suslin A.V.  Application of deep neural networks for identification of objects recorded by information-measuring system

The problem of correct identification of an object recorded by information-measuring systems during the sensing of the surrounding space is highly complex and science-intensive in its model-algorithmic resolution. Intensive development of modern artificial intelligence technologies, in particular, software solutions based on deep neural networks, creates a qualitatively new technological framework for the effective solution of related problems. The paper develops the scientific and methodological background for applying neural network models in the identification of objects recorded by an information-measuring system while carrying out a purposeful search in the geographical space.
Keywords: information-measuring systems, deep neural network, identification.

Tsuprikov A.A.  Drilling process simulation for optimal oil-well hole making control

The paper offers a new drilling process model, which describes the nature of the driving speed change as a function a the bit load. The model is based on a sine curve; its adequacy to the drilling process is demonstrated. The model can be applied in optimal control systems for oil-well drilling.
Keywords: drilling rate model, objective function, M.G. Bingham curve, model adequacy.

Larin A.A., Ushakov A.Yu.  Software-based improvement of sensor switch triggering stability under electromagnetic interference and climate factors

The paper offers a software-based method for improving the triggering stability of a sensor switch under electromagnetic interference and climate factors. The method was implemented in the MU.KS.02-00 module used in call and command touch blocks of freight/passenger elevators. Keywords: high loaded applications, flaw detection, neural networks, unstructured data, roadbed.

Monastyrsky A.V., Vlasov Yu.B.  Computer simulation system of foundry processes

Featured properties and functionality of PoligonSoft simulation system are discussed. The solution addresses the needs of foundry process engineers.
Keywords: computer simulation system, foundry processes, process engineer.

Shilov L.S., Shan'shin S.E., Romanov A.S., Shelupanov A.A.  Reconstruction of 3D model of human foot shape on the basis of video stream

The paper offers a procedure for the reconstruction of a 3D model of human foot shape using photogrammetry and machine vision. The procedure enables the reconstruction of 3D foot shape, including its low arch, with the help of a series of smartphone 2D images. Mask R-CNN neural network ensuring 97.88% accuracy is used as a model for the segmentation of foot images. The study has resulted in the reconstructed high-resolution 3D foot model. The standard deviation of length and width is 0.95 mm. The reconstructed 3D foot models can be used for 3D printing and subsequent custom footwear making, medical research, and shoes seek in Internet.
Keywords: 3D foot reconstruction, photogrammetry, machine vision, orthopedic footwear fabrication, size finding.

Zakharov N.A.  Domestic platform for business process robotization

Domestic platforms for business process robotization are overviewed. Examples of implemented robotized business tasks are cited. A new class of tools known as smart process automation is outlined.
Keywords: business process robotization, smart process automation import replacement.

Gashenko Yu.V., Gashenko A.A., Mel'nikova D.A., Astapov V.N.  The features of luminous flux distribution in the space of fiber-optic liquid medium density sensor with an open optical channel

The paper examines the luminous flux distribution in the fiber-optic sensor's space as well as the features of the control and modulation of the flow in the sensor's open channel with the help of a modulating element. The application of an attenuation type fiber-optic displacement sensor is illustrated by an example of a float densimeter where the change of liquid density is proportionate to the lattice displacement, that enables luminous flux modulation in the fiber-optic sensor's space.
Keywords: fiber-optic transducer, density, displacement, open optical channel, cutoff attenuator, optical modulation, float densimeter.