The primary concern of the present chapter is to address the distributed leader-following consensus tracking problem for a network of agents governed by uncertain diffusion PDEs with the Neumann-type boundary actuation and uncertain spatially varying diffusivity. Except for the “leader” agent that generates the reference profile to be tracked, all remaining agents, called “followers,” are required to asymptotically track the infinite-dimensional time-varying leader state. The dynamics of the follower agents are affected by smooth boundary disturbances unbounded in magnitude and with a bounded derivative. The proposed local interaction rule is developed by assuming that only neighboring collocated boundary sensing is available, and it consists of a nonlinear sliding-mode-based protocol. The performance and stability properties of the resulting infinite-dimensional networked system are then formally studied by means of Lyapunov analysis. The analysis demonstrates the global exponential stability of the resulting error boundary-value problem in a suitable Sobolev space. The effectiveness of the developed control scheme is supported by simulation results.

The paper presents developed placement algorithms that allow solving the problem of optimized placement of a given set of flat objects, taking into account additional restrictions specified when arranging real objects in cutting and packing problems. The proposed algorithms are based on the use of the author’s model of potential containers, which describes the free space used for placement in the form of a set of rectangular areas. In this case, objects of complex geometry are considered rasterized in the form of orthogonal polyhedra. A description of the algorithm for setting individual indents for all placed objects is given, restricted zones are introduced, an algorithm is proposed for uniformly arranging a set of objects in the placement space, a placement algorithm is described taking into account several starting points for placement, a speed-optimized algorithm for placing orthogonal polyhedra is proposed, and the procedure for selecting an angle of objects to be packed is described. Examples of various placement schemes obtained using the developed algorithms are given.

В данной работе исследуются показатели производительности двухкаскадного коммутатора полностью оптической сети при возникновении неисправностей и всплесков трафика. Для анализа используется модель двухкаскадной системы массового обслуживания с общим буфером, которая адекватно описывает работу полностью оптического двухкаскадного коммутатора. Получены аналитические выражения для нахождения вероятностей состояний системы, пропускной способности и размера буфера первой и второй ступеней при скачкообразном изменении скоростей поступления и обслуживания с использованием полученной матрицы трансляции вероятностей. Результаты численных экспериментов позволили сделать вывод о характере изменения показателей производительности коммутатора при возникновении неисправностей в коммутационных элементах отдельных ступеней, а также при возникновении всплесков трафика.

Reinforcement learning based methods can achieve excellent results for robot locomotion control. However, their serious disadvantage is the long agent training time and large number of parameters defining its behavior. In this paper, we propose a method that significantly reduces training time. It is based on the Policy Modulating Trajectory Generator (PMTG) architecture, which uses Central Pattern Generators (CPG) as a gait generator. We tested this approach on an OpenAI BipedalWalker-v3 environment. The paper presents the results of this algorithm , showing its effectiveness in solving a locomotion problem over challenging terrain.

The goal of this article is to analyze a repairable double hot standby system equipped with a single repair facility, employing Marked Markov Processes. It is assumed that elements’ life- and repair times follow arbitrary distributions. The proposed approach facilitates the calculation of the system’s reliability characteristics and allows investigating their sensitivity to the shape of input distributions. Building on prior research that investigated the reliability function and mean system lifetime, the current article focuses on analyzing the steady-state probabilities of the system. For this, two system repair scenarios are considered. To validate the new method, numerical examples that are compared to previously obtained analytical results, demonstrating a high degree of accuracy, are presented. Additionally, further numerical experiments illustrate the sensitivity of steady-state reliability measures to the initial parameters of the system.