Belarusian State University of Transport

The problem of bending of a five-layer circular plate symmetric in thickness by axisymmetric distributed load is considered. The central and outer layers are assumed to be load-bearing, thin, and of increased rigidity. They perceive the main part of the force load and can exhibit elastoplastic properties. Their deformation follows the Kirchhoff hypotheses. Two relatively thick nonlinearly elastic fillers are used to connect the load-bearing layers. They provide redistribution of forces between the layers and are used to protect against unwanted external influences such as temperature and radiation. The deformation of the fillers is described by Timoshenko hypotheses, which take into account the relative shift, the additional rotation of the normal to the middle surface of the layer. The system of differential equations of equilibrium for the considered plate is obtained using the variational method of Lagrange. It includes a system of two nonlinear differential equations. The sought-for functions are the deflection of the plate, the radial displacement of the midplane of the central supporting layer, and two relative shear displacements in the fillers. The Ilyushin method of elastic solutions is used to solve the corresponding boundary value problem. The general solution is obtained in a recursive form. Formulas are provided for calculating the sought-for displacements and relative shear displacements under the boundary conditions of rigid clamping of the plate contour. The convergence of the method and the dependence of the solution on the physical nonlinearity of the layer materials are numerically investigated.

The paper discusses mathematical and simulation methods for modeling audio frequency track circuits. Their limitations are shown: static nature of analytical approaches and rigid topology binding in simulation models. The necessity of transition to digital twins as a tool for dynamic analysis and diagnostics is substantiated. A concept of a digital twin for AFTC based on a verified simulation model using spectral signal analysis is proposed.

To increase the service life of fillet welded joints, it is proposed to chamfer the end face of one of the joined parts in the weld area. It is shown that the resulting reduction in the weld leg size does not impair the reliability of the welded joint. Optimal weld leg sizes are determined.

Under the traffic congestion scenario, the charging demand of electric vehicles has significant temporal and spatial fluctuations, which pose higher requirements for the operation and scheduling of optical storage and charging microgrids. Therefore, aiming at the dual uncertainties of traffic congestion and PV output, this study proposes a collaborative optimization method by integrating dynamic demand forecasting and time-sharing elastic scheduling. Firstly, an ultra-short-term prediction model of photovoltaic power generation based on Variational Mode Decomposition-Long Short-Term Memory (VMD-LSTM) and an error compensation mechanism is established to realize high-precision prediction of photovoltaic output under different weather conditions. Secondly, by combining a traffic flow dynamic model and user behavior analysis, a charging demand prediction model under congestion scenarios is constructed. On this basis, a time-sharing electricity price optimization and energy storage coordinated dispatching model is also established with the dual objectives of maximizing the total revenue of the microgrid and customer satisfaction, which is solved by the dynamic programming method. The experimental results indicate that the mean absolute percentage error (MAPE) of the proposed PV prediction model is 2.72% under various weather conditions, which is better than that of VMD-LSTM (2.84%), SVM (6.22%), and LSTM (4.76%). Under three typical congestion scenarios, the system revenue increases by about 15.8% and 18.7% respectively through dynamic adjustment of electricity price and optimal scheduling of energy storage, while the customer satisfaction only decreases by about 1%. Obviously, the proposed strategy is very effective in improving economy and maintaining good user experience. The main innovations lie in that a photovoltaic ultra-short-term forecasting model integrating VMD-LSTM and an error compensation mechanism is proposed, a charging demand forecasting model considering the dynamic response of traffic congestion is constructed, and a time-sharing rolling optimization framework is designed to realize the coordinated scheduling of photovoltaic, energy storage, and congestion.

The article discusses a method for determining the parameters of a biexponential approximation of an electrostatic discharge pulse. The work uses integral equivalence criteria: rise time, total pulse energy, and effective spectral width. A genetic algorithm is applied for multi-criteria optimization.