The paper is devoted to solving the actual problem of developing software systems for computer-aided design and program execution for the effective implementation of numerical algorithms. The theoretical basis of our research is the concept of Q-determinant. This concept suggests a unified representation of numerical algorithms in the form of Q-determinants. The Q-determinant makes it possible to express and evaluate the parallelism resource of the algorithm, as well as show the method of its parallel execution. In the paper we consider the general principles of software systems organizing for computer-aided design and program execution for the effective implementation of numerical algorithms. We demonstrate the realization of these principles by the examples of the creation of specific software systems. In addition, we compare the approaches to the implementation of the considered software systems.

For the assessment of mineral deposits and the identification of potential drilling zones using seismic data, it is necessary to build a consistent seismic stratigraphic model. The construction of such a model is carried out in several stages: fieldwork, initial data processing, calculation of transformed signals, and interpretation. Consistency in the interpretation results is achieved by adhering to fundamental stratigraphic principles. This work considers two approaches to automating the identification of reflection horizons in seismic cubes, considering prior information based on Lee algorithm. A distinctive feature of the proposed solution is its speed, low proportion of sequential execution, solution stability, consideration of constraints set by the user, and adherence to two stratigraphic principles (original horizontality and superposition). In a test seismic cube with 27951 traces, comprising 200 measurements with two user-defined horizons, the tracing of horizons was completed in 12 seconds using double the amount of RAM compared to the size of the processed data. The developed approach does not solve the problem of fault detection but can take their presence into account due to the ability to set fixed borders tracked by the user as function argument.

This paper presents a detailed comparative analysis of statistical tests utilizing both modern data compressors and standard statistical methods for assessing the randomness of Random number generators (RNG). Our study aims to thoroughly evaluate the efficiency and performance of these tests in determining the quality of RNG output sequences.
Data compression techniques have long been recognized as effective statistical tests, with some being asymptotically optimal. We compare the effectiveness of these data compressor-based tests with traditional statistical tests in assessing the randomness properties of RNG.
Our results demonstrate that the efficiency of data compressor tests and standard statistical tests is closely similar. Through rigorous experimentation and analysis, we show that both approaches yield comparable results in evaluating the randomness of RNG output sequences.

The aim of this study is to develop a corporate context-aware question-answering system in the form of a chatbot to support territorial management by providing fast and accurate access to relevant information. The system is built upon large language models leveraging the Retrieval-Augmented Generation (RAG) approach, combined with modern data processing and retrieval techniques. The knowledge base of the system incorporates textual and tabular data extracted from official reports on environmental conditions. A PostgreSQL database with the pgvector extension was employed to store and retrieve large volumes of vectorized data efficiently. Advanced Bi- and Cross-encoder architectures were integrated to enhance the accuracy of context interpretation and the relevance of retrieved answers. The system’s backend was implemented using the Text Generation Interface platform, while the frontend relies on ReactJS and a Telegram bot, ensuring an intuitive user interface for diverse audiences. Additional features, such as query classification filters and temporal interval normalization, were introduced to refine the system’s response precision and user experience. The results demonstrate the effectiveness of combining state-of-the-art artificial intelligence technologies with robust data management solutions. The developed system not only improves response accuracy and retrieval efficiency but also offers scalability and adaptability for application in various domains. This research highlights the potential of AI-driven solutions in addressing complex challenges in territorial management and beyond.

The task of primary breakup of a liquid jet in a gas flow and its subsequent complete atomization in a two-phase approximation is being solved. The carrier phase is gas, and the dispersed phase is liquid and its droplets formed as a result of breakup. The VOF (Volume of Fluid) model, based on the Euler-Euler approach, implemented using ANSYS software [1], is used for the solution. In the VOF model, the transport of each phase is described by their volume fractions—continuous functions of time and spatial variables. In the momentum conservation equation for the mixture, the interaction of liquid and gas with the mixture is described by the surface tension force, determined as a function of the curvature and normal vector to the interface between liquid and gas. The VOF model describes the primary breakup of the liquid jet. The Mixture and Eulerian models, from the same Euler-Euler group of ANSYS models, are also capable of describing the primary breakup of the liquid jet, but they require the average droplet diameter of the liquid phase (characteristic size of the dispersed phase particles) for their closure. It is impossible to use the Euler-Euler approach, based on convective-diffusion equations for concentration, mass, and momentum of particles (EECD) [2; 3], to describe the primary breakup of the liquid jet. However, EECD, with lower computational costs than the VOF model, describes atomization, starting from the complete atomization region. The complete atomization region of the liquid is required for the start of the Euler-Lagrange approach [4], in which the dispersed phase is described by tracking the trajectories of droplets throughout the computational domain. The droplet trajectories are calculated in the flow field of the carrier phase, obtained from the Navier-Stokes equations. Thus, complete atomization is needed for models with coefficients depending on the characteristic particle size of the dispersed phase. In two-step methods, the first step finds complete atomization using a model that describes primary breakup. The second step starts from complete atomization using an economical model. In the article, the VOF model is used to find the complete atomization of the liquid jet, providing the volume fraction distribution of the dispersed phase. A method is proposed for its interpretation in terms of droplets, verified and validated on problems of liquid film breakup [5; 6] and kerosene atomization in gas turbine mixing channels [7].