The first chapter focuses on information technologies, which are an important aspect of modern society, determining the development and functioning of various spheres of human activity. It describes the basis of information technology on the processing, transmission and storage of data using computers and other technical means. Attention is also drawn to the key characteristic of information technology - the speed of data processing and transmission, which has increased due to the constant development of hardware and software. Such technologies are used in various industries, including business, medicine, science and education. In addition, the role of information technology in ensuring data security, including protection against unauthorized access and cyberattacks, is considered. The concepts of the "Internet of Things" are described, where environmental objects are equipped with sensors and are able to exchange data.
The section also covers cyber-physical information technologies, which represent the integration of physical systems with information and communication technologies. This approach creates a unified ecosystem aimed at creating intelligent, autonomous systems that combine the real world with the virtual world. It is noted that cyber-physical information technologies allow you to interact with physical objects in real time, using smart algorithms, sensors and communication networks. Such technologies include the use of smart sensors, data collection from physical objects and their integration with cloud-based data processing and analysis systems. In industry, cyber-physical information technologies can be used to create "smart factories" where automation and monitoring systems interact with equipment and personnel. In medicine, these technologies can support the creation of intelligent medical systems and devices for diagnosis and treatment. In general, cyber-physical information technologies define a new level of integration between the physical and digital worlds, opening up new prospects for the development of intelligent systems and optimization of various industries.
The chapter provides a comprehensive overview of the current state of the artificial intelligence paradigm, which allows agents to learn optimal strategies through interaction with the environment. It is noted that reinforcement learning has made significant progress in recent years due to deep learning methods, increased computing power, and new algorithmic developments.
The chapter provides a rationale for the theoretical foundations of reinforcement learning, pointing out the significant achievements, challenges, and potential future directions in the field. Particular attention is paid to key concepts and cutting-edge research that informs the progress made in reinforcement learning in various fields, such as robotics, gaming, and decision-making systems.
This chapter provides a compiled and systematic summary of the current state of research in reinforcement learning, providing the reader with a comprehensive picture of the achievements, challenges, and prospects of this important area of artificial intelligence.
The second section is devoted to a detailed analysis of the environments in which the reinforcement learning task experiments were conducted. This section discusses the dependence of the system learning rate in the reinforcement learning method on the number of mutually independent modules, as well as the comparison of the capabilities of search module systems and individual objects in finding targets with unknown locations in a known environment using reinforcement learning algorithms.
The capabilities of heterogeneous UAV swarms using reinforcement learning and a decision-making system using reinforcement learning to control heterogeneous UAV swarms are also investigated. This chapter describes the environments in which the experiments were conducted, their characteristics and parameters, and the methodology and procedures for conducting the experiments.
The chapter also discusses the results of the experiments and their interpretation, which allows us to draw conclusions about the effectiveness of the applied methods and approaches in solving problems in the context of reinforcement learning in different environments and scenarios.
In the second section, we conducted a study aimed at determining the dependence of the system learning rate in the reinforcement learning method on the number of mutually independent modules in the environment. In accordance with the described parameters and conditions of the experiment, three identical environments were chosen, each with a different number of research objects: one, five, and ten, respectively. The research objects were the same in each environment.
