Vol 26, No 3 (2022)

We present a system approach to evaluating dynamic forms of interactions between elements of mechanical oscillatory systems, which can be applied as design diagrams of engineering objects used for transport or technology purposes. The approach of structural mathematical modelling was used, where a mechanical oscillatory system (with the finite number of freedom degrees) is compared to the block diagram of a dynamically equivalent automatic control system. The research object was represented by a family of mechanical oscillatory systems with the finite number of freedom degrees under the action of connected force harmonic disturbances. The main research focus was on a set of dynamic states of an engineering object, which are caused by the application of common-phase external force disturbances at various frequencies and characterised by a connection coefficient. For a family of chain mechanical oscillatory systems with two degrees of freedom, a method for interpreting a set of dynamic states in the form of oriented graphs was developed. A method for constructing a population graph of dynamic states based on the frequency response of system transfer functions was developed. Within the framework of the developed interpretation, the graphs of dynamic states can be considered as peculiar invariants that persist on sets of mechanical oscillatory system parameters. Using the methodology of structural mathematical modelling, a concept of dynamic invariants was developed, according to which the totality of dynamic invariants can be used to provide a general evaluation of the variety of dynamic states and forms of dynamic interactions between the elements of mechanical oscillatory systems. The proposed concept of dynamic invariants extends the methodology of structural mathematical modelling in relation to the problems of system analysis for ensuring the safety of engineering objects for transport and technology purposes under the conditions of connected vibrational loading.

The feasibility of applying the method of two measurements in determining the parameters of equivalent circuits of electrical network elements for the highest harmonic components is analyzed. Experiments were carried out using a MATLAB Simulink model of the common coupling point that includes a distorting load in the form of a three-phase rectifier, a nondistorting linear load and a generalized power system without distortion sources. The parameters of an equivalent circuit in the form of active bipoles, consisting of current distortion and conductivity sources, were determined using the method of two measurements of mode parameters. Modes with variations in the active and reactive power of the studied distorting load and loads in the external electrical network were considered. The results of determining the equivalent circuit parameters under 20% and more variations in the power loading were established to be unstable (400% dispersion of actual values). Therefore, these parameter values appear to be unreliable due to their dependence on the value of external load power. At the same time, the simulation of random variations in loading parameters within 10% of the initial value allowed the parameters of an equivalent circuit to be correctly determined. It was shown that the equivalent circuit of a nondistorting linear load consists solely of conductivity, while the equivalent circuit of a distorting load can contain non-zero conductivity on the considered harmonic component. Thus, according to the performed study, the method of two measurements produces the results acceptable in terms of accuracy (deviation from actual values of less than 1%) not at a single significant variation in the mode parameters, but during a continuous monitoring of small natural variations in the parameters of the electric power system. The results obtained can be used when solving the problem of online assessing the effect of loads on the quality of electricity, since the initial data for this problem include the equivalent circuit parameters.

We investigate the effect of modernizing the main condensate scheme applied at the power unit No.5, CHP-10 of Baikal Energy Company LLC, on its energy and economic efficiency using a customized simulation model. Simulation modelling was carried out in the Machine Program Building System software environment. The constructed model was verified using the results of measuring control parameters in several operating modes according to a three-stage procedure of verifying the mathematical models of complex thermal power units. We propose an original approach to modernizing the main condensate thermal scheme at the power unit under study in order to reduce the specific fuel consumption for balance-of-plant needs. The idea was to expand the main condensate scheme by including an additional sealing pump to the feedwater electric pumps of the 1Кs-20-110 unit in order to avoid the incorrect selection of condensate pumps. The study showed that the redistribution of the main condensate flow between the existing condensate pumps and the proposed 1Ks-20-110 sealing pump leads to a reduction in the specific fuel consumption for the electric energy generation to 0.32 g of fuel per kWh across the range of electrical loads from 137 to 150 mW. As a result, the net efficiency of the power unit can be raised by 0.03%. It should be noted that the proposed modernization project may reduce the consumption of electricity for balance-of-plant needs when the feedwater electric pump is switched to the standby mode during an emergency shutdown of the power unit. Based on the performed energy and economic calculations, energy savings for the period of 2019 comprise 82,653 kWh, while the total annual savings are estimated at 78,030 rubles. Thus, the conducted research demonstrates the high efficiency of applying simulation modelling in the study and optimization of existing thermal power plants.

This paper presents a modification of the analytic hierarchy process in order to increase its efficiency for a multi-criteria comparison of mixes of generating plants in local energy areas during their development. The multi-criteria problem of selecting the most effective ratio of rated capacities is considered for power plants representing a single mix of generating plants during the development of a local energy area in the Khabarovsk Krai. The energy sources are represented by thermal, solar, wind and diesel power plants. The following estimation criteria for alternative solutions were accepted: levelized cost of electricity, ecological effeciency estimation; estimation of public opinion about the consequences involved with the establishment of power plants. In order to solve the multi-criteria problem, the analytic hierarchy process (AHP) was used. When using the original AHP for the set problem, a large quantity of alternatives at the stage of pairwise comparisons were found to represent a perceptible load on a decision maker. Thus, already during the estimation of 10 alternatives according to 5 criteria, decision makers should conduct 225 pairwise comparisons, which may eventually result in an unacceptable consistency of the results. In addition, this requires a procedure accounting for the uncertainty of the decision maker's preferences. The proposed solution represents a method of forming matrices of pairwise comparisons upon criteria. This method consists in generating an interval or fuzzy model of the decision maker's preferences for evaluating pairs of estimates according to the criterion. The proposed method was verified using a numerical example of solving the set problem. The obtained optimum mix of power plants consists of thermal, solar and diesel plants with a power of 30, 35 and 39 mW, respectively. The proposed method ensures a high consistency of the results obtained during alternative pairwise comparisons. In addition, the modified analytic hierarchy process takes into account the non-linear nature of the decision maker's preferences for estimating alternatives according to criteria.

This study aims to improve the accuracy of forecasting the electricity consumption of an enterprise based on an analysis and preliminary processing of input data, as well as at evaluating the effect caused by feature selection on the results of various forecast models. A woodworking enterprise located in Nizhniy Novgorod was selected as a forecast object. Two types of machine learning methods, including neural network and ensemble models, were compared. An approach to selecting the most significant parameters (features) from a time series was considered in order to improve the results of the following ensemble models based on decision trees: adaptive busting (AdaBoost), Gradient Boosting and Random Forest. The most significant features of the initial time series were determined using the calculation of correlation coefficients between the values of electricity consumption in forecasted and previous hours. For the considered forecast object, the most significant features were established to be the consumed energy in hours lagging behind the forecasted hour by the multiple number of days. The schedule of repair works for woodworking machines was used as an additional feature. According to the obtained results, decision tree ensembles can surpass artificial neural networks provided that significant features are selected correctly. Thus, the smallest average error of a neural network model on a test sample comprised 7.0%, while an error of 5.5% was obtained for a Gradient Boosting ensemble model. The use of a repair schedule was demonstrated to additionally increase the forecast accuracy: for the considered ensemble models, the error reduced from 20 to 30%.

This study is aimed at improving the efficiency of monitoring and automated control over the technological state of electrolytic cells during copper refining using the recognition and identification of short circuits between the electrodes. Experimental works were performed on a laboratory bench consisting of two cells connected in series to the electrical circuit. The method of detecting short circuits is based on measuring infrared radiation using a scanning thermal imager (Оptris PI 400i / PI 450i) for determining areas with elevated temperatures of electrode and electrolyte surfaces. A short-circuit detection algorithm was developed and tested in MATLAB environment using the Image Processing Toolbox MATLAB functions. The proposed method for recognizing and identifying short circuits between the electrodes of an electrolytic cell is based on using a scanning digital thermal imager. This method allows a comprehensive assessment of the contact area, determination of both the start time of a short circuit and the degree of electrode heating in the short-circuit zone, as well as establishment of temperature values in all areas, including the electrolyte temperature in the entire volume of the cell. An algorithm for searching for short-circuit locations in an electrolytic cell is proposed. This algorithm involves the collection of data and its comparison for adjusting the inter-electrode distance, ensuring a stable energy mode for the entire electrolytic line, and determining the exact position of each cathode in the cell using the threshold processing of infrared images obtained by a thermal imager. It was established that timely identification of short circuit locations in the electrolytic bath (since the onset of a contact between the electrodes), as well as recording of the growth rates of a dendritic coalescence area and electrolyte temperature, ensure timely elimination of technological disturbances during the operation of electrolytic baths. Thus, the developed method of short circuit identification for comprehensive determination and recognition of the technological state of an electrolytic cell during the electrolytic refining of copper can be used to reach a stable energy mode with minimum deviations in the process temperature.

In this work, we develop a digital twin of a tube digester and a software application automatically calculating the aluminate solution quality based on mass balance equations with the purpose of improving the quality of hydro-chemical process management at the Rusal Achinsk JSC alumina plant. A mimic diagram of the nepheline sinter leaching process was visualized using the Wonderware InTouch software package. Quick scripts were created to display the process animation. The input parameters include the solution temperature, the consumption of sinter and recycled solution, as well as the chemical composition of raw materials. The air temperature, vibrations and equipment malfunction were used as the main disturbing effects. Output parameters consisted of sludge and aluminate solution yields, power consumed by the drive and the silicon ratio of the resulting solution. Based on the mass balance of sinter and recycled solution, calculated using industrial laboratory data, an algorithm for modelling variations in the digestion process and a software application for calculating the quality of aluminate solution were developed using the InTouch built-in capabilities of scripting in its own programming language. A digital twin of a tube digester was developed on the basis of an operating unit utilized at the Rusal Achinsk JSC. It was shown that the developed software application predicts the quality of the finished product (aluminate solution) using a simple balance model under variations in both the composition of raw materials (cake and recycling solution) and the rate of fed raw materials, as well as present the results of calculations performed using a mimic diagram of the process. Thus, the developed interface can be used to simulate various technological operations performed in a tube digestor, including solution drainage, sludge discharge and variations in the inlet flows. At the same time, it becomes possible both to observe current changes in process parameters and archive plots, as well as to select the optimal composition of raw materials to obtain an aluminate solution of higher quality.

Viktor Elshin, Doctor of Technical Sciences, Professor, Head of the Department of Automation and Control of the Irkutsk National Research Technical University celebrates two anniversaries on October 3, 2022: his 75th birth anniversary and the 50th anniversary of his scientific and pedagogical activity.