Volgograd State Agricultural University (Russian: Волгоградский государственный аграрный университет) is a public university located in Volgograd, Russia.
Statement of the problem. The aim of the work is to apply the constitutive equations of the theory of plastic flow obtained by the authors without separating the strain increments into elastic and plastic parts to determine the stress-strain state of structures in which stress concentration zones exceeding the yield strength of the material are formed under loading. Results. The proposed version of the constitutive equations of the theory of plastic flow without separating the strain increments into elastic and plastic parts is used to obtain the stiffness matrix of a mixed prismatic finite element with nodal unknowns in the form of displacement increments and stress increments at the loading step. The sought values of the internal point of a finite element with triangular bases were approximated using linear functions. Conclusions. A specific example shows the practical coincidence in the results of calculations using the constitutive equations of the theory of flow and the proposed version of the theory of plasticity.
The article discusses the problems of automated identification of various classes of landscape sections based on their color images. Solving the problem of multi-ple classification requires substantiating the architecture and hyperparameters of deep neural networks, their parameterization and deployment on mobile devices, taking into account the features of the analyzed objects. The purpose of the study is to substantiate the neural network architecture for mobile cyberphysical systems for automated detection of color image areas. The quality and performance of the projected neural network significantly depend on the characteristics of the train-ing dataset, therefore, various methods of preprocessing the source images are used, including image dimensionality correction. Well-known neural network architectures, including fully connected, convolutional, and recurrent layers and their combinations, were analyzed. An analysis and justification of the basic ar-chitecture of the neural network were carried out to identify the characteristic are-as of color images. To eliminate the influence of stochasticity of the initial choice of its weights, a deterministic approach is justified and mathematical dependen-cies are proposed for choosing the initial values of the weights of a neural net-work. The neural network quality values were obtained: average accuracy in the training sample: 0.958; maximum accuracy in the training sample: 0.967; average accuracy in the test sample: 0.912; maximum accuracy in the test sample: 0.921.
The perspectives on the sustainable development of the transport and logistics complex include the need to minimise the environmental impact and increase social needs satisfaction. Transport and logistics are the main components of the country’s socioeconomic development (Sumbal et al., 2023). The key to success in the logistics sector lies in cooperation between humans and machines. Sustainable transport in logistics involves implementing practices and technologies that substantially reduce the environmental impact of transport and distribution activities. Sustainable transport aims to minimise greenhouse gas emissions, reduce air and water pollution, and optimise resource use. Integration of innovations and sustainable development is important for the optimisation of logistic operations and the growth of effectiveness and support of companies in achieving the Sustainable Development Goals. A successful transition to more sustainable logistics requires such approaches as the integration of eco-friendly technologies, an increase in renewable energy use, and the promotion of digital solutions to raise effectiveness. Technological progress, combined with the development of environmental awareness and regulatory requirements, stimulates the fundamental evolution in this sector. Quality 5.0 in the transport and logistics sector combines leading technologies, such as artificial intelligence, machine learning, and blockchain, with an approach that is oriented towards humans and sustainable development. The new approach puts emphasis on cooperation between humans and machines to create more effective and sustainable supply chains. Quick technological changes and the increasing needs of consumers are the main drivers of digital transformation. Logistics and transport companies have to accelerate the implementation of models that are based on data analysis, to use new market opportunities to the largest extent.
