Journal of Computer and System Sciences International

Adequate mathematical modeling of various dynamic control processes leads to control problems with nonlocal conditions, which are described by both partial differential equations and ordinary differential equations. Control problems with non-local conditions arise, on the one hand, as mathematical models of real processes, and on the other hand, as a result of the inability of formulating problems correctly using local conditions. Non-local problems in the form of integral conditions, in particular occur in the mathematical modeling of phenomena of various natures. Research into problems with non-local conditions, in various formulations, is important not only for theoretical purposes but also for practical necessity. This paper investigates the problem of controlling a linear dynamic system subject to a non-local integral-type condition. The control problem for a linear dynamic system is formulated with an integral condition imposed on the components of the phase vector over a certain portion of the time interval during the system’s operation. Conditions are obtained, under which a solution to the system of ordinary linear differential equations with the given integral condition exists, and this solution is constructed. Control functions are developed for the control problem, under the influence of which the phase trajectories are realized, satisfying the integral conditions of the problem. The continuity and non-uniqueness of the control functions are demonstrated. the criterion of complete controllability of a linear system with phase constraints of integral type is obtained. An example is given as an illustration.

The problem of rejection of linear-bounded exogenous disturbances, i. e. those growing no faster than a linear function of the system state, is considered for linear control systems. The quality is characterized by the size of the bounding ellipsoid containing the system output. A regular approach to solving this problem is proposed; it reduces the initial problem to an equivalent parameterized semidefinite programming, easily solved numerically. It is based on the technique of linear matrix inequalities and the method of invariant ellipsoids. The efficiency of the proposed procedure is demonstrated by a test example.

The paper is devoted to the review of nonlinear filtering algorithms based on Monte Carlo methods. The general principles of their design, advantages and disadvantages are discussed. The main focus is on recursive algorithms, which use sequential importance sampling and resampling procedures. The application of the considered algorithms in processing of navigation data is demonstrated.

A four-wheel-drive mechanical system with individual torque sources for each wheel is considered. The study addresses the problem of optimal control of a vehicle whose initial state is characterized by a substantial lateral velocity. The objective is to design a control law that ensures safe maneuver execution: preventing collision with a stationary lateral obstacle while simultaneously achieving the maximum possible velocity in a prescribed direction. The problem is solved numerically using Pontryagin's Maximum Principle.

The problem of job scheduling in a multiprocessor real-time computing system is solved for the case when requests for task execution are received cyclically with specified periods. In addition to processors, there are also non-renewable resources. The volumes of job execution are assumed to be fixed. An algorithm for optimal distribution of non-renewable resources and construction of an admissible schedule is developed. The algorithm is based on reducing the original problem to finding a flow in a special type of network. In the case where a solution does not exist, it is shown how the periods of receipt of requests should be increased so that a solution exists. The issue of stability of solutions was investigated.

A multi-criteria algorithm for constructing a flight route bypassing dangerous zones is presented. The following three characteristics are considered as criteria for evaluating optimality: the time spent in the danger zone, the secrecy of the route and the length of the route. The article describes a method for constructing a set of alternative solutions, as well as various algorithms for choosing a rational solution.

The problem of optimal program control of spatial motion (in particular, maneuvering) of a spacecraft considered as a free rigid body of arbitrary dynamic configuration with a quadratic functional of the energy spent on spacecraft maneuver and a fixed transition time is investigated. In the class of helical generalized conical motions an optimal analytical solution of the problem is obtained under arbitrary boundary conditions of angular and linear positions and angular and linear velocities of spacecraft, which is brought to the algorithm. Four–dimensional dual Euler (Rodrigue-Hamilton) parameters, which are components of the biquaternion of the finite Clifford displacement, are used to describe spatial motion. The solution is obtained using the Chasles theorem on the displacement of a free rigid body and the Kotelnikov–Studi displacement principle. Numerical examples showing the effectiveness of the proposed solution to the problem are given.

The controlled rolling of a rimless wheel up a slope with a rod attached to the center of the wheel is investigated. The control is carried out by deflecting the rod by the electric drive from the support spoke. The optimal rod control in the single-bearing phase of motion for wheel acceleration is found. It is shown that there is a value for the angle of deviation of the rod from the vertical, for which this angle is not perturbed upon impact due to the change of the support spokes, and this angle is energetically optimal for movement as a whole. The Poincare sequence and diagram for the angular velocity of rotation of the wheel step by step are constructed and the conditions for the existence of stationary points of the diagram are revealed.

This paper examines the application and development of technologies for quantitatively assessing human–machine interfaces intended for use in controlling a group of robots. This article presents an analysis of existing technologies for evaluating human-machine interfaces. The analysis reveals that currently used technologies do not fully support a comprehensive quantitative assessment of human-machine interfaces, necessitating the development of appropriate formal tools and computational instruments. For the development of technologies for evaluating human-machine interfaces (commander–operator – robot group), this approach is proposed. This approach utilizes computational procedures to determine quantitative indicators of the compliance of the actual characteristics of the interfaces being evaluated with the required ones. A distinctive feature of the proposed approach is its comprehensive nature, which allows for the formalization of the most significant, key qualitative characteristics using expert methods to obtain sufficiently reliable and valid interface assessments, expressed as numerical values. The proposed approach, in essence, can be applied to obtain a quantitative assessment of a wide range of human-machine interfaces, as well as to evaluate other information and communication tools for human interaction with technical systems when creating fundamentally new human-machine (ergatic) systems.