No 3 (2023)

A locomotion system is considered in the form of a chain of a finite number of bodies (materi along points) moving in a straight line on a horizontal rough plane due to the forces of interaction between the bodies. These forces serve as the control variables. Dry Coulomb friction acts between the bodies and the plane. The necessary and sufficient conditions are obtained under which the nonreversible motion of all bodies of the system for the same distance is possible under the assumption that in the initial and final positions the velocities of all bodies are equal to zero. Nonreversible motion is understood as a motion in which none of the bodies changes the direction of their velocity in the process of moving.

With the development of data collection technologies, a significant volume of multiview data has appeared, and their clustering has become topical. Most methods of multiview clustering assume that all views are fully observable. However, in many cases this is not the case. Several tensor methods have been proposed to deal with incomplete multiview data. However, the traditional tensor norm is computationally expensive, and such methods generally cannot handle undersampling and imbalances of various views. A new method for clustering incomplete multiview data is proposed. A new tensor norm is defined to reconstruct the connectivity graph, and the graphs are regularized to a consistent low-dimensional representation of patterns. The weights are then iteratively updated for each view. Compared to the existing ones, the proposed method not only determines the consistency between views but also obtains a low-dimensional representation of the samples using the resulting projection matrix. An efficient optimization algorithm based on the method of indefinite Lagrange multipliers is developed for the solution. The experimental results on four data sets demonstrate the effectiveness of the method.

Methods for optimizing project schedules for the criterion of minimizing the weighted average time of their execution are considered. In the case when the durations of the jobs are given deterministically, an exact and approximate method for solving the problem of choosing the optimal schedule is proposed. If changes in the durations of the jobs are possible, an analytical tool for estimating the stability of the schedules is created both for the situation of interval setting of the durations of the jobs and for the situation of changes in the durations of the jobs under possible disturbances in the external environment. In the event that the durations of the jobs are given stochastically, a mechanism for evaluating the effectiveness of the schedule by two criteria is proposed, and a procedure for the quantitative assessment of the risk of the schedule is proposed.

The functionals of dynamic systems of various types are optimized using modern methods of reinforcement learning. The linear resource allocation problem, as well as the optimal consumption problem and its stochastic modifications are considered. In the reinforcement learning strategy gradient methods are used.

A method for determining the values of the attitude angles of an aerial vehicle (AV) is considered. For the proposed method, the initial data are the values of overloads in the body-fixed coordinate system and velocity projections in the Earth’s normal coordinate system. It is proposed, using the approaches of direct methods for finding the optimal control, to parametrize the pitch, roll, and yaw angles, and then find the values of the parameters based on the initial data. The resulting optimization problem can be solved using a population algorithm.

This article studies the development of new effective methods for designing spacecraft (SC) trajectories for missions using the generalization of the Rutherford formula for the scattering of charged particles proposed by the authors for the case of gravitational scattering. For controlled gravity assist maneuvers of SC, a rule is formulated that makes it possible to purposefully create beams of trajectories with specified properties as a result of a gravity assist maneuver. Modifications of this rule are presented that make it possible to implement efficient and reliable beam recurrent procedures for searching for the ballistic scenarios of interplanetary flights. Of particular importance is the synthesis of sequences of gravity assist maneuvers that provide the given change in the asymptotic velocity of the SC relative to the target planet.

The motion control of the Nauka Multipurpose Laboratory Module (MLM) during its flight to the International Space Station (ISS) is considered. For maneuvering and angular stabilization, a propulsion system is used as an actuator. In order to ensure simultaneous control of the motion of the center of mass of the spacecraft (SC) and its stabilization with the help of engines at each moment of time, it is necessary to solve the problems of determining the required change in the speed of the SC, choosing the optimal configuration of the engines, and the problem of predicting the motion parameters of the SC. Methods for solving these problems, applied in the development of the control system of the Multipurpose Laboratory Module, are presented. The operability of the described algorithms is confirmed by the results of mathematical modeling on a ground test bench for onboard software and in the course of flight design tests directly during the flight and successful docking on the ISS.

A robotic complex consisting of a transport-handling platform and a climbing robot installed on it is considered. The platform is designed to move the robot on a horizontal surface and place it on a vertical surface. The practical implementation of the platform control algorithm is described, which makes it possible to automate the process of placing the robot on a vertical surface.