#12 2019

CONTENTS №12

Zakharov N.A.  5G and Industry 4.0

Featured properties of 5G network technology important for industry are discussed. Key implementation features of 5G networks are listed with the focus on millimeter-wave communications. The possibility network structure sharing owing to network slicing, network functions virtualization, and software-defined networking is shown.

Keywords: 5G, IoT, Industry 4.0, network slicing, network functions virtualization, software-defined networking, millimeter wavelength range.

Discussing a Topic…

Artificial intelligence for industrial sites

Yan D.E.  Artificial intelligence as a company employee

The paper examines two available digital transformation technologies: robotic process automation (RPA) decreasing human involvement in repetitive tasks, and predictive behavioral analytics. Both enable automatic discovering of organization’s best practices. Each of the two helps mankind to understand itself better and organizations to work more effectively.

Keywords: artificial intelligence, automation of repetitive tasks, human potential analytics, digital trace, predictive behavioral analytics.

Koshevoy V.O., Vishnevsky K.V., Pronchenkov I.A., Chwernyshova E.A., Salakhov I.I., Zurbashev A.V.  Comparison of predictive NIR spectroscopy techniques for oil product quality analysis

Application of NIR spectroscopy coupled with mathematical modeling is relevant for evaluating and predicting the properties of hydrocarbon petroleum compounds with various compositions. A partial least square (PLS) regression and an ANN model were built for predicting physicochemical properties of Diesel fuels. The models are compared with respect to prediction effectiveness and development potential. ANN applicability for NIRS-based chemometric analysis is shown.

Keywords: NIR spectroscopy, Diesel fuels, chemometrics, regression analysis, PLS regression techniques, artificial neuron networks.

Lifanov A.V., Sokolov D.I.  ANN-based artificial intelligence for industrial sites

The paper outlines the ANN concept and discusses underlying hardware platforms. It presents ANN-based Siemens solutions and includes implementation cases.

Keywords: artificial neuron networks, cloud platforms, ANN learning, neuroprocessor module.

Yukhno A.I., Plugotarenko N.K.  Application of intelligent data processing techniques for chlorine agent metering automation

Intelligent data processing techniques are applied for chlorine agent metering based on drinking water quality assessment. A neuron network model is used for predicting the content of organochlorine compounds generating during water chlorination. Public health risks associated with water disinfection are assessed, threshold values of organochlorines content are calculated subject to cancer risk assessment. Fuzzy water quality indices were developed during the research using Mamdani fuzzy inference system. A water chlorination algorithm ensuring safe water consumption is developed on the basis of trapezoidal membership functions and applied for chlorine agent metering and maintaining water quality.

Keywords: metering, chlorination, cancer risk, fuzzy logic, prediction.

Glazyrin A.E.  Intellectualization of immersive environment for a training simulator

The paper examines the possibility of improving process operators training efficiency with immersive environment (IE) simulators. A two-level procedure of intelligent IE design is offered. At the lower level, the simplest operator’s actions are selected along with special operator productivity assessment test, while the efficiency indicators and IE parameters affecting the complexity of task execution are to be chosen at the upper one. The procedure presumes the availability of “result – task parameters adjustment” feedback in the training system. An algorithm generating an individual operator training program is developed, which allows for the values of operator’s actions efficiency indicators. The algorithm enables the development of an individual training program and takes into account operator’s productivity based on laboriousness parameters changes. Experimental results have confirmed the expedience of the algorithm application for technological transport simulators.

Keywords: immersive environment, training simulator, fuzzy model, training, operator, genetic algorithm.

Kunchinin A.N., Krikunov D.E.  How artificial intelligence can increase oil extraction at mature oil fields

Electric centrifugal pumps are widely applied in oil industry. Decisions on pump mode changes are traditionally made by engineers to ensure best performance characteristics, such as shaft speed, operation time cycle, liquid extraction to accumulation ratio for the cycle. The paper offers a machine learning based solution for the optimal control of oil extraction mode. The pilot solution was implemented at an oil filed in West Siberia.

Keywords: electric centrifugal pumps, artificial intelligence, machine learning, optimal control, oil extraction.

Veresnikov G.S., Pankova L.A., Pronina V.A.  Software toolkit for parametric synthesis tasks in the design of technical facilities under parameter uncertainties

The paper discusses optimization tasks of parametric synthesis in the design of technical facilities under parameter uncertainties with the help of soft computing. In the context of uncertainty theory, it offers a method for modeling optimizable parameters epistemic uncertainty. Optimization models where objective functions and constraints depend upon input and optimizable uncertain parameters are developed and investigated. A modeling technique for the functions dependent on uncertain and random parameters is developed. Based on the theoretical results, a pilot software toolkit is developed.

Keywords: uncertainty theory, mixed uncertainty, software tool, parametric synthesis, design, genetic algorithm.

Acharya M.P., Mousavi M.J.  Real-time artificial intelligence systems in digital twins on peripheral devices

Cloud platform based digital twins are currently widely used and deliver consider benefits. At the same time, placing a pre-trained in a cloud digital twin on a peripheral device is a novel application of artificial intelligence enabling real-time industrial data processing outside the cloud. Solutions advantages are demonstrated; application examples are cited.

Keywords: artificial intelligence, cloud technologies, digital twin, modeling, peripheral devices, real time.

Vladov R.A., Dozortsev V.M., Shaidullin R.A., Belousov O.Yu.  Predictive analytics in chemical industry

The paper analyzes the features of big data generated by chemical plants. These features determine the possibility and the ways of applying state-of-the-art monitoring and predictive analytics tools for process equipment. Honeywell Forge for Industrial is the first-of-its-kind production analytics platform for big data processing big data processing the analysis of process and equipment condition and offering the recommendations for intelligent process control, reliable and optimal operations. Data acquisition and preparation, key concepts of equipment modeling, predictive detection and processing of events related with process upsets and equipment failures are examined. Featured properties of various predictive analytics methods are discussed, and application case is included.

Keywords: big data, chemical processes, condition-based maintenance, equipment status monitoring, predictive analytics, diagnostic and prognostic models, pattern recognition, mean time between failures.

Rotchev K.V.  Rosemount 5300 radar meters for reliable interphase level measuring under possible gas phase occurrence

Operation principle of interphase level meters is discussed. The techniques of interphase level measuring in fully or partially filled process tanks with Rosemount 3300 and 5300 wave-guide radar level meters are described. Level meter’s operation mode is selected dependent on its application conditions (entire or partial sinking of the wave guide).

Keywords: wave-guide radar level meters, interphase changes, measured medium.

Potapov A.I., Shikhov A.I., Dunaeva E.N.  Temperature monitoring of permanently frozen soils

The effect of temperature on the properties of permanently frozen soils is shown, methods and tools for their monitoring are outlined. The paper notes that the problem of permanently frozen soils monitoring is not yet fully resolved.

Keywords: permanently frozen soils, temperature monitoring, cryolithozone, thermometric well, logger, thermometric scythe.

Morozenko A.A., Zyablov A.A., Kovalenko J.A.  Design concepts of a complex knowledge base of standard institutional and managerial solutions with reference A-plant construction

The problem of knowledge deployment in planning and management of production processes in the field of A-plant construction are discussed against the lack a unified approach to the development of project management system, to lessons learned application, to the search and identification of the necessary information from the completed projects, as long as the lack of a unified information storage format. The need for creating the methodology of automated search of standard institutional and managerial solutions and developing the appropriate knowledge base to underlie a unified information toolkit is substantiated. The toolkit will combine the existing heterogeneous knowledge storage systems, identify unambiguously and replicate the accumulated experience of A-plant power unit construction.

Keywords: knowledge base of standard solutions, standard institutional and managerial solutions, intelligent information system, A-plant.

Journal’s Club

Salikhov Z.G.  Multifunction automated complex for ferrous, nonferrous, and municipal trash neutralization

The paper discusses the problem of industrial and household hazardous waste disposal and offers dual-zone Vanyukov furnaceы for this purpose for their obvious advantages against other present-day technologies. It mentions considerable application experience of such furnaces in Russia coupled with their long-term scientific research.

Keywords: Vanyukov furnace, municipal solid waste disposal, autogenous smelter.