An approach to managing the main modes of smelting steel in heavy-duty electric arc furnaces (EAF) using digital twin technology was defined and formulated. It was noted, that the existing power regulators do not have the function of balancing the effective power of phases and, accordingly, electric arcs because they are focused on working with a certain average value of the signal. It is proposed to use the analysis of dynamic characteristics based on instantaneous values of input parameters instead of operating ones, as it’s usually implemented in most devices. This allows us to obtain more accurate data on the arc state and reduce the amount of time and computing power required to obtain a result and form recommendations. Based on the data obtained as a result of long - term observations of the heavy-duty EAF-135 operation, the relationship of the constant component of the arc voltage (CCAV) with the metal oxidation is shown. An example of its use as a criterion for controlling the melting oxidative stage is given. This reduces the consumption of electrochemical sensors for each melting in the case of serial metal production. Based on the recorded data, it is possible to timely determine the unevenness of the arc power release between the furnace electrodes and issue recommendations on gas burners operation regulating to equalize the rate of scrap melting at electrodes with less power release. The authors propose the idea of using digital twins based on models of the active power distribution across the melting bath zones and dependence of metal oxidation on oxygen blowing for monitoring and controlling the electric mode and the oxygen blast mode at the oxidative stage of the melting process. Simplified schemes of these twins are given.

This paper considers the use of machine learning for diagnosis of diseases that is based on the analysis of a complete gene expression profile. This distinguishes our study from other approaches that require a preliminary step of finding a limited number of relevant genes (tens or hundreds of genes). We conducted experiments with complete genetic expression profiles (20 531 genes) that we obtained after processing transcriptomes of 801 patients with known oncologic diagnoses (oncology of the lung, kidneys, breast, prostate, and colon). Using the indextron (instant learning index system) for a new purpose, i.e., for complete expression profile processing, provided diagnostic accuracy that is 99.75% in agreement with the results of histological verification.

This paper considers the use of machine learning for diagnosis of diseases that is based on the analysis of a complete gene expression profile. This distinguishes our study from other approaches that require a preliminary step of finding a limited number of relevant genes (tens or hundreds of genes). We conducted experiments with complete genetic expression profiles (20 531 genes) that we obtained after processing transcriptomes of 801 patients with known oncologic diagnoses (oncology of the lung, kidneys, breast, prostate, and colon). Using the indextron (instant learning index system) for a new purpose, i.e., for complete expression profile processing, provided diagnostic accuracy that is 99.75% in agreement with the results of histological verification.

This work presents the results of observing the oxidation stage of steelmaking in EAF-135 with the help of an automated control system of electrical characteristics. The workspace of the furnace has three main zones which differ by the physical state of materials in them: arc discharge, melt and foamed slag. The data on the distribution of electric power over the furnace zones are given. It is shown that the main factors impacting the melt and slag resistances are oxygen blowing and motion of electrodes. The influence of magnesia flux feeds on the melt resistance is studied. It is noticed that these feeds correspond to a sharp increase and subsequent gradual decrease in the resistance, whereas the dope congestion time does not exceed one minute. The average electrical parameters of the workspace zones are given for single heats in EAF-135. The nature of the change in the arc discharge power and the change in the melt temperature are compared. It is noted that the profiles of changes in these characteristics match one another, and the rise in the arc power corresponds to the rise in the melt temperature. An attempt to correlate the FeO content in slag with the arc power has not given a positive result. It is noted that the control of this parameter by changes in the electrical parameters of the arc and the slag zone due to the overwhelming influence of intense oxygen blast, melt mixing and electrode motion does not meet the reliability criterion. However, this method should be checked at the refining of steel in the ladle-furnace unit.

Industrial implementation of the method of smelting industrial silicon with some fraction of the reducing agent replaced with silicon carbide is exemplified. The prospects for using this method to increase the efficiency of the reduction process and the furnace output are shown. The specific consumption of electricity and the amount of greenhouse gases decrease. Additives have a beneficial effect on the electrical efficiency of the furnace by stabilizing it, thus reducing the number of process control errors.

The results of observation of steelmaking oxidation stage at EAF-135 were analyzed with automated control system of the electrical characteristics. There are three main zones in the furnace working space, which differ by the aggregate state of the materials in them: arc discharge, melt and foamed slag. The distribution data of electric power over the furnace zones is given. There is active powers asymmetry of the arcs under single electrodes affected by asymmetry of the short network. It is shown that the main factors impacting the melt and slag resistances are oxygen blow and electrode movement. The authors studied the influence of magnesia flux feeds on the melt resistance. These feeds correspond to a sharp increase and a subsequent gradual decrease in resistance, and the time for assimilation of additives does not exceed one minute. The average electrical parameters of the working space zones are given for the EAF-135 at single heats. A comparison was made of the nature of change in the arc discharge power and the change in the melt temperature. The profiles match of changes in these characteristics to an increase in the arc power corresponds to increase in the melt temperature. An attempt to correlate FeO content in the slag with the arc power did not give a positive result. However, this methodology should be tested under conditions of steel refining in a ladle-furnace unit. It is noted that the parameter control at changing of the electrical parameters of the arc and slag zones due to the overwhelming influence of intense oxygen blast, melt mixing and electrode displacement does not meet the reliability criterion.

One of the main disadvantages of artificial neural networks is slow learning associated with the need to calculate a large number of coefficients. The article shows that learning can be substantially accelerated. Acceleration is achieved by a sharp reduction in the number of training patterns. In addition, the inverse pattern method was used both for the formation of features and for the subsequent recognition of objects; this has made it possible to dispense with coefficients, which significantly reduces the amount of calculations. In instantaneous learning, as in deep learning, features are generated automatically. Computational experiments have shown the invariance of the method with respect to not only scaling and rotations but also large deformations of objects to be recognized.

A method for controlling dynamic current-voltage characteristics in an electric arc furnace is described. The influence of the electrode’s current direct component on the accuracy of estimating the resistance of the furnace zones and the direct components of the arc current and voltage is noted. An improved algorithm for determining these parameters is presented with the results of its testing in a laboratory experiment on the use of a metallized additive to a charge for steel smelting. Not only was the addition of scale shown to have no negative effect on the process, it resulted in a slight decrease in the magnitude of the arc direct component, as well as contributing to the slag foaming process.

The annual increase in cargo turnover to the Far Eastern seaports contributes to the optimization of the railway infrastructure in order to increase the level of throughput and processing capacity of the Baikal-Amur Mainline and Trans-Siberian Railway facilities. JSC Russian Railways pays special attention to these issues. Organization and promotion of car traffic through the formation of freight trains using the connection technology, “virtual coupling”, heavy and dual trains that move on an ongoing basis is one of the innovative methods for increasing the carrying capacity for highly loaded sections and directions at a level of utilization of their throughput of 0, 8 or more. Under these conditions, the organization of efficient power supply of railway sections and facilities is the main factor influencing the value of the throughput capacity of transport infrastructure facilities. The article presents a technical justification for the measure aimed at increasing the throughput and processing capacity of a section of a railway line, taking into account the stability check of the devices of the traction power supply system. The predicted parameters of the operation of the traction power supply system of the railway line section are given with a promising train schedule.

The authors analyzed data concerning the content of the starting materials and products of the intermediate reactions of silicon reduction, as well as electrical parameters of the workspace zones in the ore-reduction furnace baths. It was noted that the state of the technological process and its parameters largely depend on the content of silicon carbide in the burden and melt zones and its relationship with the silicon monoxide flow intensity. Examples illustrating negative effect of the corrective reducing additives on the technological process and furnaces performance are provided.

On the basis of data on the distribution of electric power in the working space of a furnace, resistances of the zones of charge, arc, and melt, and the characteristics of displacements of electrodes obtained with the help of an automated system of monitoring of the electric parameters of the process, we estimate the efficiency of the electric mode of a two-electrode single-phase furnace used for smelting commercial silicon. The influence of the methods and procedures used for the regulation of the conditions of charge dosing and landing of electrodes on the electric mode is investigated. It is shown that the requirements of stabilization of the useful power and its rational distribution over the zones imposed on the regulation of the electric mode are not satisfied. The instability of the useful power of the furnace is caused by the application of the manual mode of control and the errors of measurements of currents in the electrodes caused by the incorrect calibration of the reference ammeter. The realization of the technological process under the conditions of high landing of electrodes does not correspond to the rational values of the energy efficiency of the furnace and is accompanied by significant losses of the arc power of the furnace and silicon in the form of silicon monoxide. The absence of data on the state of the working space of the furnace does not enable us to control the distribution of power over its main zones and to realize the possibility of full control over the process. We present examples that confirm the oscillating nature of the technological process and significant asymmetry of the useful power of separated electrodes. It is shown that numerous violations in the course of the analyzed process are caused by the use of irrational methods and procedures of regulation of the modes of charge dosing and landing of electrodes. In particular, the alternation of the modes of loading of the charge with excess and lack of carbon leads to significant fluctuations in the useful power of the furnace and the powers of the arcs. Moreover, the course of smelting can be described as a permanent transient process from one extremely inefficient state into another.

This study describes the conversion processes of charge materials and the chemical processes in the charge and hearth zones of furnaces that smelt industrial silicon and high-silicon ferroalloys. The primary factor that determines the efficiency of the charge mode and the technological process is the coefficient of excess carbon in the furnace bath. The fluctuations in the effective capacity of the furnace; its distribution over the zones of the arc, charge, and melt; and the active resistances of these zones are caused by the regular deviations of this coefficient from unity. Such deviations are induced by uncontrolled changes in the humidity of the reducing agents. Carbon deficiency and excess modes affect the state of charge and melt zones and the distribution of electric energy on the arc discharge power. Transient processes that occur when the composition of the charge on the dosing unit is changed, as well as the direct supply of the reducing agent to the furnace mouth and reflecting changes in the charge zone conductivity, are studied. Results reveal that transient processes are accompanied by high material and energy losses and their duration, depending on the volume of the bath and the effective capacity of the furnace, can exceed 4 h. On the basis of the analysis of the changes in charge conductivity and arc power, the negative consequences of the layer-by-layer charge loading method adopted for smelting industrial silicon are presented; these consequences are based on the alternation of charge supply, with an excess and shortage of the reducing agent. In the absence of reliable methods for the continuous monitoring of the moisture content of reducing agents in the charge stream, the coefficient of carbon excess must be evaluated by changing the resistance of the melt zone, which reflects changes in the size of the carbide layer that is formed due to the imbalance of silicon carbide and silicon monoxide. Such evaluation can determine the degree of imbalance of the reducing agent in the mixture and apply the currently required control effect on the composition of the charging material in a timely manner.

The mathematical processing results of a number of EAF-135 melts electrical parameters are presented. The influence of the foamy slag formation, molten metal carburization and the harmful impurities oxidation on the change in electrical parameters is analyzed. The methodology for determining the metal oxidation from the electrical melt parameters is characterized. In the presented form, the technique is cumbersome for the rapid assessment of the oxidation process intensity and requires refinement. Keywords: Metal oxidation, electric parameters, electric arc furnace

Modern silicon production technology is associated with a risk of negative environmental impact due to the fact that in addition to the final product, other reaction products are formed, including dust, from the incomplete use of charge materials. Gases released during silicon smelting in ore-thermal furnaces are characterized by the content of a large amount of fine dust. Dust consists of 94–96% of silicon dioxide. As a result of the use of sulfur-containing raw materials in furnaces as sulfur reducing agents, sulfur compounds in the form of SO2 are present in the furnace gases entering for purification, and nitrogen oxides are also present. The developed silicon recovery smelting technology reduces the technological energy consumption and increases the furnace productivity in proportion to the amount of carbon replaced by silicon carbide. Replacing carbon with silicon carbide reduces the dust content and the amount of exhaust furnace gases, and changes their composition. Thus, reducing the amount of pollutants reduces their anthropogenic impact on the environment. Keywords: silicon, gas cleaning dust, gas capture system, microsilica

Silicon carbide is obtained in ore-thermal furnaces by reduction of silica (quartzite) with carbon. The use of silicon carbide in the production of technical silicon as a carrier of the target element and as a reducing agent can significantly improve the technical and economic performance (TEP) melting. The process of reducing silicon melting in electric furnaces takes place in two stages. First, silicon carbide is formed as a pseudomorphosis over the carbon of the reducing agent, then silicon carbide interacts with silicon oxide to form elementary silicon. Physical and chemical properties of silicon carbides obtained with the use of various reducing agents were studied. The reducing potential and reaction ability of carbides depends on how their surface is developed. Carbide volume and density characteristics are obtained on the matrices of charcoal and petroleum coke. For comparison, data for carbide obtained in the Acheson furnace are presented. Measurements of relative electrical resistivity of the reducing agent were performed and obtained on the carbides basis with temperature in the range of 700–1700∘C. For comparison, the RER values of silicon carbide obtained in the Acheson furnace are given, the resistance of carbides is several times higher than the RER of the corresponding reducing agents, which favorably affects the furnaces smelting silicon electric mode. As a result of the silicon carbide addition to the charge, the power of the arc discharge increases and the intensity of the reduction process increases. Keywords: silicon carbide, gas cleaning dust, gas capture system