Модель та метод інформаційної технології навігації автономних мобільних систем в умовах невизначеності

Abstract

UA: Об'єкт дослідження – процес навігації автономних мобільних систем. Проведені дослідження ґрунтуються на використанні методів ситуаційного управління, нечіткої логіки, теоретико-множинного опису, математичної логіки, теорії експерименту і комп'ютерного моделювання, машинного навчання з підкріпленням, теорії оптимізації, функціонального моделювання. У результаті дослідження вирішено наукове завдання з розробки моделі та методу інформаційної технології навігації автономних мобільних систем, що забезпечує однорідну інтеграцію таких чотирьох завдань навігаційного циклу, як сприйняття інформації, локалізація, динамічне планування маршруту та ситуаційне керування рухом у недетермінованому частково спостережуваному оточенні. Сферою застосування результатів дисертації є системи моніторингу розосереджених об'єктів та інші системи, що включають як елемент, автономні мобільні системи.

EN: The thesis is devoted to the solution of an important scientific problem on the development of a model and method of the information technology of navigation of an autonomous mobile system (AMS), which provides homogeneous integration of four navigation cycle tasks such as information perception, localization, dynamic path planning and situational motion control in an indeterminate partially observed environment. Parts of the research, the results of which are presented in the thesis, were performed within the research work of the Ukrainian State University of Railway Transport "Smart machines and the Internet of Things with artificial intelligence", № DR 0120U104276 and the state-funded research work of the Kharkiv National University of Radio Electronics "Deep hybrid systems of computational intelligence for data streams analysis and their fast learning", № DR 0119U001403. The research object is the process of navigation of autonomous mobile systems. The research subject encompasses models, methods and technologies of navigation of autonomous mobile systems in the conditions of uncertainty, incompleteness and fuzzies of information. The research aim is to expand the possibilities of navigation of mobile systems in the autonomous mode in a partially indeterminate environment. To achieve the research aim the following research tasks have been solved: 1) an analysis of existing models and methods of navigation used in autonomous mobile systems has been carried out; 2) a model of situational control of movement on a path of an autonomous mobile system in the conditions of the provided disturbances based on the perception conceptual model with mechanisms of context management and attention switching has been developed; 3) a model and method of situational control of movement on a path with dynamic re-planning of a path of an autonomous mobile system in the conditions of unforeseen disturbances have been developed; 4) the information technology of navigation of an autonomous mobile system based on the model of situational control of movement and dynamic re-planning has been developed; 5) comprehensive testing of the developed model, method and information technology has been carried out. The scientific novelty of the thesis is as follows:  for the first time, a model of situational control of an autonomous mobile system movement with re-planning of a path has been developed, which, unlike others, takes into account the impact of unforeseen disturbances that make it impossible to move to the goal position of the path, and due to situational re-planning based on sensory data about the current situation allows an autonomous mobile system to achieve the goal;  the context-sensitive method of situational control has been improved, which is adapted to the task of controlling the movement of an autonomous mobile system along a given path, which allows on this basis to solve the problem of navigation of mobile autonomous systems in conditions of uncertainty and incomplete information;  the further development received the information technology of navigation of an autonomous mobile system using the model and method of situational control with re-planning of the path, which allows to expand the use autonomous mobile systems in a partially indeterminate environment, when there is no possibility or economically unreasonable a streamlining of the environment.