The work addresses the problem of simulating grasping an object of a specific shape by a robotic manipulator. The robot under consideration is similar to the human hand in terms of kinematic configuration. The modeling mechanism includes several stages: positioning the robotic hand over the object, placing the manipulator’s fingers relative to the shape of the object grasped, calculating the contact force required to grasp the object, and moving the hand holding the object. The first and last phases are conducted by changing the coordinates of the robotic manipulator’s kinematic center. The finger positioning stage is reduced to solving the inverse kinematics problem through an evolutionary optimization method, i.e. a genetic algorithm. The grasping phase defines the relevant contact force based on resolving the tracking issue for a set value, which is a predetermined touch force. The amount of the assigned touch force is reliant on the required friction force between fingers and the object that ensures a firm grasp. The authors approximate the finger dynamics to second-order dynamics. The simulation findings reveal the applicability of the proposed method to the problem of object grasping using a robotic manipulator.

The article studies implementation of the structured query language (SQL) in systematizing data on vehicles, their owners, and road traffic accidents (RTAs) within an integrated information environment. The paper aims to develop and validate a method for designing a relational database that ensures accounting integrity and financially precise automated calculation of insurance payments. The work structures table with vehicle registration plates data using DDL scripts, which include constraints for data compliance with state standards. The authors perform the key business logic with an SQL query that applies conditional logic and aggregation for the accurate accounting of liability limits under the compulsory motor third party liability insurance. The study contains a comparison of the relational approach with non-relational (NoSQL) and scalable relational (NewSQL) methods. It is concluded that standard SQL systems are most efficient for accounting tasks requiring strong consistency with ACID properties and calculation transparency.

The paper analyses modern topologies of high-frequency isolated direct current (DC) converters, including phase-controlled bridge circuits and resonant LLC converters. The study defines the following key disadvantages of the aforementioned converters: significant circulating power, narrow soft switching range, and increased switching losses in a number of modes. The scope of the research encompasses a dual-transformer resonant topology, which provides advantages in distributing energy flows and suppressing peak currents. Simulation of the converter's steady-state operation using the fundamental approximation has shown that the output power can be substantively manipulated by means of a single characteristic, i.e. phase-shift of the half-bridges control signals. The authors study the transformer ratio's influence on the limit conditions of zero voltage switching (ZVS). The article establishes that the optimal ratio of transformer inductances significantly expands the load range at which soft switching of all power switches remains unchanged up to the open-circuit. The work develops a converter design method that contains an algorithm for optimizing the power contour characteristics to minimize conduction losses and expand the ZVS range. The technique consists in a sequential selection of the leakage inductance, transformer ratio, magnetic core shape and number of windings' turns, ensuring compliance with the set requirements in terms of efficiency, weight and size. The experiment confirms high energy efficiency of the converter (over 96%) in a wide load range and stable maintenance of the soft switching mode.

The paper models the dynamics of connectivity change in irregular networks using percolation theory. The article analyzes an existent knowledge network, i.e. citation system for scientific publications based on a sample from the arXiv database related to the "second superstring revolution" in high-energy physics (HEP). The selected numerical modeling method enables a closer study of the network clustering and the macroscopic quantities' values near the percolation threshold. The work demonstrates that the percolation threshold increases with the average network degree, reaching a saturation value of 0.65–0.70 in the node problem. Moreover, the average cluster size also increases. The growth rate of node clustering peaks in the range from 0.5 to 0.8 considering a single node omission probability. The results reveal that clustering in an expanding knowledge network is more sensitive to changes in the connection density compared to percolation processes.

This article presents an algorithm for generating and improving patent portfolios via genetic programming and particle swarm optimization. The specified algorithm enables selection of an optimization method according to the following parameters: data organization, expected results interpretability, resources availability, and task size. The study defines genetic programming as a recommended method for small portfolios in solving strategic tasks that require solution interpretability, while particle swarm optimization as applicable practice for rapid optimization of large portfolios. The author demonstrates the algorithm's efficiency in aircraft engineering by addressing existent problems using self-created software.

Visualization of a Russian-language ontology using the pydot library in Python leads to the creation of a knowledge graph with incorrectly displayed Cyrillic characters. Assignment of a Russian-language supporting font while defining the graph has no influence on the final image. The study proposes a solution to this problem which lies in setting an explicit encoding when exporting the image as a file via the .write_png('output.png', encoding='utf8') command. This procedure enables the creation of a knowledge graph with Cyrillic script support. The introduced method of Russian-language ontologies visualization in Python can be applied by IT specialists in developing information systems based on ontologies under consideration or by educators using the ontologies in the teaching process.

The article considers the problem of multi-criteria decision analysis under uncertainty conditions caused by subjective weights of the criteria. The paper addresses the specified issue as well as proposes an approach of interval criteria estimation instead of point assessment. This technique improves the model's sufficiency and solutions stability. The study presents a modification of the TOPSIS method, which lies in its adaptation for interval-defined weights on an infinite set of alternatives. Moreover, the approach introduces a soft TOPSIS concept that is resistant to changes in criteria weights within set limits. The author tests the method's practical significance and efficiency by optimizing a multicomponent mixture formula.

The paper considers calculating the heating and cooling of a two-stage rotating cylinder in a dry granulation unit for liquid blast furnace slag. In order to address the issue, the authors develop a mathematical model and an algorithm for estimating the temperature field of the two-stage rotating cylinder upon intense end heating and lateral surface cooling. The study proposes the implementation of the numerical scheme with set initial data in computer optimization of the cylinder's heat exchange parameters when designing a shaft disk atomizer of the aforementioned unit.

The work addresses the problem of identifying a human fatigue state experienced during work activities. The authors develop software for exhaustion recognition and subsequent voice notification using pyttsx3 library. Conducted testing of the designed unit reveals the possibility of its implementation in people’s fatigue detection.

The paper proposes an optimization method of the Walsh–Hadamard codes, the augmented Walsh code and the p-ary Walsh–Hadamard code, via matrix encryption by a code matrix. The study shows that the upgrade does not affect the process of error detection and correction. Decoding algorithms using code matrices are developed as an addition to list decoding. In the p-ary case, the authors demonstrate the efficiency of the optimization method in handling errors of different types such as substitution, erasure, and appearance of a symbol.